R/querygenescrb.R
In alexcis95/PhenoExam: Compare the likeness of phenotypes terms between groups of genes
Defines functions
forbes
CGIloaddercompareOpt
UNIPROTloaddercompareOpt
PSYGENETloaddercompareOpt
ORPHANETloaddercompareOpt
CTDloaddercompareOpt
GENOMICS_ENGLANDloaddercompareOpt
CLINGENloaddercompareOpt
CRBloaddercompareOpt
MGDloaddercompareOpt
HPOloaddercompareOpt
CGIloaddercompare
UNIPROTloaddercompare
PSYGENETloaddercompare
ORPHANETloaddercompare
CTDloaddercompare
GENOMICS_ENGLANDloaddercompare
CLINGENloaddercompare
CRBloaddercompare
MGDloaddercompare
HPOloaddercompare
CGIloadderOpt
UNIPROTloadderOpt
PSYGENETloadderOpt
ORPHANETloadderOpt
CTDloadderOpt
GENOMICS_ENGLANDloadderOpt
CLINGENloadderOpt
CRBloadderOpt
MGDloadderOpt
HPOloadderOpt
UNIPROTloadder
CGIloadder
PSYGENETloadder
ORPHANETloadder
CTDloadder
GENOMICS_ENGLANDloadder
CLINGENloadder
CRBloadder
MGDloadder
HPOloadder
PhenoEnrichGenespransimu
PhenoGeneNumberSimuok
PhenoRelation
foundgeneoverlap
PhenoEnrichCompareRandom
PhenoEnrichGenespran
PhenoEnrichCompare
entrezmap
convertgenesymbolmessage
PhenoEnrichRandomOpt
PhenoEnrichGenes
RandomComparePheno
PhenoGeneNumberOpt
PhenoGeneNumber
FindPhenoFromGenes
ListGenesPheno
findoverlapgenes
getdbnames
QueryGenes
Documented in
CGIloadder
CGIloaddercompare
CGIloaddercompareOpt
CGIloadderOpt
CLINGENloadder
CLINGENloaddercompare
CLINGENloaddercompareOpt
CLINGENloadderOpt
convertgenesymbolmessage
CRBloadder
CRBloaddercompare
CRBloaddercompareOpt
CRBloadderOpt
CTDloadder
CTDloaddercompare
CTDloaddercompareOpt
CTDloadderOpt
entrezmap
findoverlapgenes
FindPhenoFromGenes
forbes
foundgeneoverlap
GENOMICS_ENGLANDloadder
GENOMICS_ENGLANDloaddercompare
GENOMICS_ENGLANDloaddercompareOpt
GENOMICS_ENGLANDloadderOpt
getdbnames
HPOloadder
HPOloaddercompare
HPOloaddercompareOpt
HPOloadderOpt
ListGenesPheno
MGDloadder
MGDloaddercompare
MGDloaddercompareOpt
MGDloadderOpt
ORPHANETloadder
ORPHANETloaddercompare
ORPHANETloaddercompareOpt
ORPHANETloadderOpt
PhenoEnrichCompare
PhenoEnrichCompareRandom
PhenoEnrichGenes
PhenoEnrichGenespran
PhenoEnrichGenespransimu
PhenoEnrichRandomOpt
PhenoGeneNumber
PhenoGeneNumberOpt
PhenoGeneNumberSimuok
PhenoRelation
PSYGENETloadder
PSYGENETloaddercompare
PSYGENETloaddercompareOpt
PSYGENETloadderOpt
QueryGenes
RandomComparePheno
UNIPROTloadder
UNIPROTloaddercompare
UNIPROTloaddercompareOpt
UNIPROTloadderOpt
#' @title QueryGenes
#' @description It produces a data frame with the phenotypical terms from the genes selected and their entrez ids.
#' @param genes A vector with the genes symbol
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes symbol and "mouse" if you have mouse genes symbol
#' @details This function has developed by Alejandro Cisterna Garcia as part of his PhD mentored by Juan Antonio Botia Blaya
#' @import readr
#' @import data.table
#' @import ggplot2
#' @import stats
#' @import plotly
#' @import ggpubr
#' @import dplyr
#' @import viridis
#' @import clusterProfiler
#' @import parallel
#' @import purrr
#' @import DT
#' @import Hmisc
#' @import pheatmap
#' @export
#'
QueryGenes = function(genes, database = "HPO", organism = "human"){
if (organism == "human") {
if (database == "UNIPROT") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CGI") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "PSYGENET") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "ORPHANET") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CTD") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "GENOMICS_ENGLAND") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CLINGEN") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "DIS") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "HPO") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CRB") {
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(custom, genes, by = "entrez")
}
if (database == "MGD" | database == "MOUSEDB") {
# Read mouse data base
musedata <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"))
names(musedata) <- c("entrez","symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info", "source" )
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his MP
onlyg <- inner_join(musedata, genes, by = "entrez")
onlyg <- onlyg[,c(1,2,5,6)]
}
}
if (organism == "mouse") {
# genes = musedata[1:50,3]
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = unique(onlygenes)
if (database == "UNIPROT") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CGI") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "PSYGENET") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "ORPHANET") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CTD") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "GENOMICS_ENGLAND") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CLINGEN") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "DIS") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "HPO") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CRB") {
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(custom, genes, by = "entrez")
}
if (database == "MGD" | database == "MOUSEDB") {
# Read mouse data base
musedata <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"))
names(musedata) <- c("entrez","symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info", "source" )
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his MP
onlyg <- inner_join(musedata, genes, by = "entrez")
onlyg <- onlyg[,c(1,2,5,6)]
}
}
return(onlyg)
}
#' @title getdbnames
#' @description This function show the databases available in PhenoExam package.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
getdbnames = function(){
sources <- c("HPO", "CRB", "MGD", "CLINGEN",
"GENOMICS_ENGLAND", "CTD", "ORPHANET",
"PSYGENET", "CGI", "UNIPROT")
return(sources)
}
#' @title findoverlapgenes
#' @description This function get the genes that are linked to the phenotypes selected.
#' @param phenos A vector with the phenotypes id that you want.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
findoverlapgenes = function(phenos){
# Un archivo base de datos con todas las DB
sdb <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/sdb.csv"), header = T)
# Make an empty df with names
rbjointri = data.frame(matrix(vector(), 0, 5,
dimnames=list(c(), c("entrez", "symbol", "term_name", "term_id", "source"))),
stringsAsFactors=F)
aparece <- length(phenos)
# For each database name in a vector run its loadder
for (ph in phenos) {
fenos = dplyr::filter(sdb, sdb$term_id == ph)
# join data frames generated
rbjointri <- rbind(rbjointri, fenos)
}
genesaparecen <- rbjointri %>% dplyr::group_by(symbol,entrez) %>% dplyr::tally() %>%
dplyr::filter(n >= aparece)
genesoverlap <- unique(genesaparecen$symbol)
return(genesoverlap)
}
#' @title ListGenesPheno
#' @description This function make a list with genes selected and his phenotypical terms
#' @param genes A vector with the genes symbol
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes symbol and "mouse" if you have mouse genes symbol
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
ListGenesPheno = function(genes, database = "HPO", organism = "human"){
# We will keep only with the genes in the query and his HP
onlyg <- QueryGenes(genes, database, organism)
# We make a list with genes and his HP
gbygen <- split(onlyg$term_id , onlyg$symbol)
return(gbygen)
}
#' @title FindPhenoFromGenes
#' @description Get the genes selected and only their phenotypical terms defined by the user.
#' @param genes A vector with genes symbol.
#' @param phenoid A vector with id terms to get.
#' @param database A string parameter with different options: "HPO", "DIS", "CRB" and "MGD".
#' @param organism A string parameter: "human" if you have human genes symbol and "mouse" if you have mouse genes symbol
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
FindPhenoFromGenes = function(genes, phenoid, database = "HPO", organism = "human"){
# We will keep only with the genes in the query and his phenotypic terms
onlyg <- QueryGenes(genes,database, organism )
# Convert vector to a dataframe
phenoid <- as.data.frame(phenoid)
names(phenoid) <- "term_id"
# We only keep genes and particular phenotypic term
onlygenesandpehno <- inner_join(onlyg, phenoid, by = "term_id")
return(onlygenesandpehno)
}
#' @title PhenoGeneNumber
#' @description This function counts the number of genes per phenotype term from the vector of genes.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param genes A vector with human gene symbol.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoGeneNumber = function(genes, database= "HPO", organism = "human"){
if (organism == "human") {
if (database == "UNIPROT") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CGI") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "PSYGENET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "ORPHANET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CTD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "GENOMICS_ENGLAND") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CLINGEN") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "DIS") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "HPO") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CRB") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "MGD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "human_symbol"
genes$HumanSymbol <- as.character(genes$human_symbol)
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$human_symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info" , "source" )
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name ) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
}
if (organism == "mouse") {
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = unique(onlygenes)
if (database == "UNIPROT") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CGI") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "PSYGENET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "ORPHANET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CTD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "GENOMICS_ENGLAND") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CLINGEN") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "DIS") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "HPO") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CRB") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "MGD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "human_symbol"
genes$HumanSymbol <- as.character(genes$human_symbol)
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$human_symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info" , "source" )
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name ) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
}
return(numportergenes)
}
#' @title PhenoGeneNumberOpt
#' @description This function counts the number of genes per phenotype term from the vector of genes.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param genes A vector with human gene symbol.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoGeneNumberOpt = function(genes, database= "HPO", organism = "human"){
if (organism == "human") {
if (database == "UNIPROT") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CGI") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "PSYGENET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "ORPHANET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CTD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "GENOMICS_ENGLAND") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CLINGEN") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "DIS") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "HPO") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CRB") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "MGD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info" , "source" )
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name ) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
}
if (organism == "mouse") {
# genes = musedata[1:50,3]
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = unique(onlygenes)
if (database == "UNIPROT") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CGI") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "PSYGENET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "ORPHANET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CTD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "GENOMICS_ENGLAND") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CLINGEN") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "DIS") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "HPO") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
names(custom) <- c("entrez","symbol", "term_id", "term_name")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CRB") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
names(custom) <- c("entrez","symbol", "term_name", "term_id")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "MGD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "human_symbol"
genes$HumanSymbol <- as.character(genes$human_symbol)
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$human_symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info" , "source" )
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name ) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
}
return(numportergenes)
}
#' @title RandomComparePheno
#' @description This function measures statistically significant phenotype similarities between gene sets and detects significant differential phenotypes for them.
#' @param geneset A vector with genes symbol.
#' @param genesetcompare A vector with gene symbol.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param p.value Cut of p value as they are relevant terms
#' @param nulltestnumber A number of random genes subset.
#' @param seed A seed.
#' @param url A logical parameter. T if you want interactive links
#' @param linux A logical parameter. T if you use linux SO, this is the default option
#' @param adjust A string parameter. You have two options: "bonferroni" is the default option and you can choose "fdr" for False Discovery Rate.
#' @param POR A string parameter. You have two options: Use the Jaccard index "jaccard" is the default option and you can choose the Forbes coefficient "forbes".
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
RandomComparePheno = function(geneset, genesetcompare, organism = "human", database = "HPO", p.value = 0.05, nulltestnumber = 100, seed = 1, conditional_case = F, url = F, linux=T, adjust = "bonferroni", POR = "jaccard"){
if (conditional_case == F) {
# Get the Phenotype Enrichment Analysis from geneset
referencia = PhenoEnrichGenes(genes=geneset, database, organism, url = url)
enrichresult = as.data.frame(referencia$htmltabla)
# Obtain the significative phenotypes terms from geneset
if (adjust == "bonferroni") {
relevantes <- dplyr::filter(enrichresult, enrichresult$bonferroni_pvalue <= p.value)
}
if (adjust == "fdr") {
relevantes <- dplyr::filter(enrichresult, enrichresult$FDR_adjust <= p.value)
}
# Get the Phenotype Enrichment Analysis from genesetcompare
targetenrich = PhenoEnrichGenes(genesetcompare, database, organism, url = url)
targetenrich = as.data.frame(targetenrich$htmltabla)
nphenotargetall <- nrow(targetenrich)
# Obtain the significative phenotypes terms from genesetcompare
if (adjust == "bonferroni") {
targetrelevant <- dplyr::filter(targetenrich, targetenrich$bonferroni_pvalue <= p.value)
}
if (adjust == "fdr") {
targetrelevant <- dplyr::filter(targetenrich, targetenrich$FDR_adjust <= p.value)
}
nphenotarget <- nrow(targetrelevant)
# Obtain overlap
phenounion <- full_join(enrichresult, targetenrich, by = "term_id")
prunionn <- full_join(enrichresult,targetenrich, by = c("term_id","term_name","source"))
nphenounion <- as.numeric(nrow(phenounion))
# Inner phenotypes
phenopresentall <- inner_join(enrichresult, targetenrich, by = "term_id") # Innner de fenotipos totales entre los dos datase
ncompartidosall <- nrow(phenopresentall) # Número de fenotipos totales compartidos
# RPOR interesccion entre union
if (POR=="jaccard"){
scoreshared <- ncompartidosall/nphenounion
}
if (POR=="forbes"){
scoreshared <- forbes(enrichresult$term_id, targetenrich$term_id)
}
# Obtain union sifnificative
if (adjust == "bonferroni") {
set1join <- dplyr::filter(prunionn, prunionn$bonferroni_pvalue.x <= p.value)
}
if (adjust == "fdr") {
set1join <- dplyr::filter(prunionn, prunionn$FDR_adjust.x <= p.value)
}
set1join <- set1join[,c(1:5,7:12,14:17)]
names(set1join) <- c("term_id", "term_name","source", "bonferroni_pvalue_set1", "FDR_adjust_set1", "gene_overlap_set1", "overlap_ratio_set1", "pvalue_set1", "common_genes_set1","bonferroni_pvalue_set2","FDR_adjust_set2", "gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
# Prepare for differentional analysis
set1join$gene_overlap_set1[is.na(set1join$gene_overlap_set1)] <- 0
set1join$gene_overlap_set2[is.na(set1join$gene_overlap_set2)] <- 0
set1join[is.na(set1join)] <- 1
# Obtain differentional phenotypes from geneset
if (adjust == "bonferroni") {
tabledif1u <- dplyr::filter(set1join, set1join$bonferroni_pvalue_set2 > p.value)
# The same for genesetcomaper
set2join <- dplyr::filter(prunionn, prunionn$bonferroni_pvalue.y <= p.value)
}
if (adjust == "fdr") {
tabledif1u <- dplyr::filter(set1join, set1join$FDR_adjust_set2 > p.value)
# The same for genesetcomaper
set2join <- dplyr::filter(prunionn, prunionn$FDR_adjust.y <= p.value)
}
set2join <- set2join[,c(1:5,7:12,14:17)]
names(set2join) <- c("term_id", "term_name","source", "bonferroni_pvalue_set1", "FDR_adjust_set1", "gene_overlap_set1", "overlap_ratio_set1", "pvalue_set1", "common_genes_set1","bonferroni_pvalue_set2","FDR_adjust_set2", "gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
set2join$gene_overlap_set1[is.na(set2join$gene_overlap_set1)] <- 0
set2join$gene_overlap_set2[is.na(set2join$gene_overlap_set2)] <- 0
set2join[is.na(set2join)] <- 1
if (adjust == "bonferroni") {
tabledif2u <- dplyr::filter(set2join, set2join$bonferroni_pvalue_set1 > p.value)
}
if (adjust == "fdr") {
tabledif2u <- dplyr::filter(set2join, set2join$FDR_adjust_set1 > p.value)
}
# Order differentional phenotypes tables
tabledif2u <- tabledif2u[order(tabledif2u$bonferroni_pvalue_set2),]
tabledif1u <- tabledif1u[order(tabledif1u$bonferroni_pvalue_set1),]
# Get fusion df term
setfusion <- rbind(set1join,set2join)
setfusion <- unique(setfusion)
# get and comapre relevant phenotypes terms
relevantesunion <- full_join(relevantes,targetrelevant, by = "term_id")
runionn <- full_join(relevantes,targetrelevant, by = c("term_id","term_name", "source"))
nrelevantesunion <- as.numeric(nrow(relevantesunion))
relevantpresent <- inner_join(relevantes,targetrelevant, by = "term_id")
ncompartidosr <- nrow(relevantpresent)
# Obtain POR
if (POR=="jaccard"){
scoreenrich <- ncompartidosr/nrelevantesunion
}
# FALTA TOCAR ESTE
if (POR=="forbes"){
scoreenrich <- forbes(relevantes$term_id, targetrelevant$term_id)
}
# Inner with all phenotypes
targetenrichttr <- targetenrich[,c(1:5,7:10)]
names(targetenrichttr) <- c("term_id", "term_name", "source", "bonferroni_pvalue_set2", "FDR_adjust_set2","gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
# Table with phenotypes
triucecompare <- merge.data.frame(enrichresult,targetenrichttr, by = c("term_id","term_name", "source"))
if (adjust == "bonferroni") {
# Label if the phenotype is relevant in both datasets
triucecompare <- triucecompare %>%
mutate(shared_enrichment = ifelse(bonferroni_pvalue_set2 <= p.value & bonferroni_pvalue <= p.value, "Yes", "No"))
# All shared phenotypes
triucecompare <- arrange(triucecompare, desc(shared_enrichment), bonferroni_pvalue_set2)
}
if (adjust == "fdr") {
# Label if the phenotype is relevant in both datasets
triucecompare <- triucecompare %>%
mutate(shared_enrichment = ifelse(FDR_adjust_set2 <= p.value & FDR_adjust <= p.value, "Yes", "No"))
# All shared phenotypes
triucecompare <- arrange(triucecompare, desc(shared_enrichment), FDR_adjust_set2)
}
targetrelevanttr <- targetrelevant[,c(1:5,7:10)]
names(targetrelevanttr) <- c("term_id", "term_name", "source","bonferroni_pvalue_set2","FDR_adjust_set2","gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
triucefinal <- merge.data.frame(relevantes,targetrelevanttr, by = c("term_id","term_name", "source"))
triucefinal <- triucefinal[order(triucefinal$bonferroni_pvalue_set),]
nspor <- nrow(triucefinal)
# Regression analysis
regresiontarget <- stats::lm(gene_overlap_set1 ~ gene_overlap_set2, data = setfusion, na.action=na.exclude)
ortarget <- summary(regresiontarget)
# Preparation for simulation analysis
# We need all genes
genAll <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/geneproteinlist.txt"))
genAll <- as.data.frame(genAll)
names(genAll) <- c("symbol","entrez")
# We transform into a vector of entrez numbers
genAll <- as.vector(genAll$entrez)
# set seed
set.seed(seed)
# To split simulation analysis
nulltestnumbernue = nulltestnumber/2
#We will keep a number of a random set of genes with the same number of genes as the genes defined
genesRandom <- replicate(nulltestnumbernue, sample(genAll, length(genesetcompare), replace=TRUE))
genesRandomg2 <- replicate(nulltestnumbernue, sample(genAll, length(geneset), replace=TRUE))
# Detect cores and work with a half of them
nworkers <- (detectCores()/2)
# Parallel simulation analysis
if (linux==T){
cl <- makeForkCluster(nworkers)
clusterEvalQ(cl, library(PhenoExam))
enrichrandomlist <- parApply(cl, genesRandom, 2, function(x) PhenoEnrichRandomOpt(x,database,organism="human"))
enrichrandomlistg2 <- parApply(cl, genesRandomg2, 2, function(x) PhenoEnrichRandomOpt(x,database,organism="human"))
stopCluster(cl)
}else{ enrichrandomlist <- apply(genesRandom, 2, function(x) PhenoEnrichRandomOpt(x,database,organism="human"))
enrichrandomlistg2 <- apply(genesRandomg2, 2, function(x) PhenoEnrichRandomOpt(x,database,organism="human"))}
if (adjust == "bonferroni") {
randomenrichrelevant <- lapply(enrichrandomlist, dplyr::filter, bonferroni_pvalue <= p.value)
randomenrichrelevantg2 <- lapply(enrichrandomlistg2, dplyr::filter, bonferroni_pvalue <= p.value)
}
if (adjust == "fdr") {
randomenrichrelevant <- lapply(enrichrandomlist, dplyr::filter, FDR_adjust <= p.value)
randomenrichrelevantg2 <- lapply(enrichrandomlistg2, dplyr::filter, FDR_adjust <= p.value)
}
# Intersección RPOR
phenopresent <- lapply(enrichrandomlist, inner_join, x = enrichresult, by = "term_name")
ncompartidos <- lapply(phenopresent, nrow)
phenopresentg2 <- lapply(enrichrandomlistg2, inner_join, x = targetenrich, by = "term_name")
ncompartidosg2 <- lapply(phenopresentg2, nrow)
# Unión RPOR
phenopresentunion <- lapply(enrichrandomlist, full_join, x = enrichresult, by = "term_name")
ncompartidosunion <- lapply(phenopresentunion, nrow)
phenopresentuniong2 <- lapply(enrichrandomlistg2, full_join, x = targetenrich, by = "term_name")
ncompartidosuniong2 <- lapply(phenopresentuniong2, nrow)
SharedScore <- mapply(function(x,y) (x/y), x = ncompartidos, y = ncompartidosunion, SIMPLIFY = F)
SharedScoreg2 <- mapply(function(x,y) (x/y), x = ncompartidosg2, y = ncompartidosuniong2, SIMPLIFY = F)
# Aqui tengo que adaptarlo al forbes
if (POR=="jaccard"){
SharedScore <- c(SharedScore, SharedScoreg2)
}
if (POR=="forbes"){
relev1 <- lapply(enrichrandomlist, function(x,y) {
forbes(x$term_id,y)
}, y = enrichresult$term_id)
relev2 <-
lapply(enrichrandomlistg2, function(x,y) {
forbes(x$term_id,y)
}, y = targetenrich$term_id)
SharedScore <- c(relev1, relev2)
}
# Intersección POR
colnames <- c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust","gene_associated_in_db", "gene_overlap", "overlap_ratio", "pvalue")
randomenrichrelevant <- lapply(randomenrichrelevant, setNames, colnames)
relevantpresent <- lapply(randomenrichrelevant, inner_join, x = relevantes, by = "term_name")
ncompartidosr <- lapply(relevantpresent, nrow)
randomenrichrelevantg2 <- lapply(randomenrichrelevantg2, setNames, colnames)
relevantpresentg2 <- lapply(randomenrichrelevantg2, inner_join, x = targetrelevant, by = "term_name")
ncompartidosrg2 <- lapply(relevantpresentg2, nrow)
# Unión POR
relevantpresentunion <- lapply(randomenrichrelevant, full_join, x = relevantes, by = "term_name")
ncompartidosrunion <- lapply(relevantpresentunion, nrow)
relevantpresentuniong2 <- lapply(randomenrichrelevantg2, full_join, x = targetrelevant, by = "term_name")
ncompartidosruniong2 <- lapply(relevantpresentuniong2, nrow)
EnrichScore <- mapply(function(x,y) (x/y), x = ncompartidosr, y = ncompartidosrunion, SIMPLIFY = F)
EnrichScoreg2 <- mapply(function(x,y) (x/y), x = ncompartidosrg2, y = ncompartidosruniong2, SIMPLIFY = F)
if (POR=="jaccard"){
EnrichScore <- c(EnrichScore, EnrichScoreg2)
}
if (POR=="forbes"){
porrelev1 <- lapply(randomenrichrelevant, function(x,y) {
forbes(x$term_id,y)
}, y = relevantes$term_id)
porrelev2 <-
lapply(randomenrichrelevantg2, function(x,y) {
forbes(x$term_id,y)
}, y = targetrelevant$term_id)
EnrichScore <- c(porrelev1, porrelev2)
}
}
if (conditional_case == T) {
# Obtener el análisis de enriquecimiento del dataset de referencia
referencia = PhenoEnrichGenes(genes=geneset, database, organism, url = url)
enrichresult = as.data.frame(referencia$htmltabla)
# Obtain the significative phenotypes terms from geneset
if (adjust == "bonferroni") {
relevantes <- dplyr::filter(enrichresult, enrichresult$bonferroni_pvalue <= p.value)
}
if (adjust == "fdr") {
relevantes <- dplyr::filter(enrichresult, enrichresult$FDR_adjust <= p.value)
}
# Todos los genes de referencia en los que no exista overlap con el dataset comparado deben salir
# de la ecuación
targetenrich = PhenoEnrichCompare(geneset, genesetcompare, database, organism, url = url)
targetenrich = as.data.frame(targetenrich$htmltabla)
nphenotargetall <- nrow(targetenrich)
# names(targetenrich) <- c("term_id", "term_name", "bonferroni_pvalue","gene_associated_in_db", "total_size_ratio", "gene_overlap", "input_size_ratio", "input_total_fold", "overlap_ratio", "pvalue", "common_genes_set2")
# Tomar los fenotipos estadisticamente significativos
if (adjust == "bonferroni") {
targetrelevant <- dplyr::filter(targetenrich, targetenrich$bonferroni_pvalue <= p.value)
}
if (adjust == "fdr") {
targetrelevant <- dplyr::filter(targetenrich, targetenrich$FDR_adjust <= p.value)
}
nphenotarget <- nrow(targetrelevant)
# Pregunta 1 overlap de términos aunque estos no sean relevantes
phenounion <- full_join(enrichresult, targetenrich, by = "term_id")
prunionn <- full_join(enrichresult,targetenrich, by = c("term_id","term_name", "source"))
nphenounion <- as.numeric(nrow(phenounion))
#nphenounion <- unique(phenounion$term_id)
#nphenounion <- length(nphenounion) # Número de fenotipos totales UNICOS DE LA UNION
phenopresentall <- inner_join(enrichresult, targetenrich, by = "term_id") # Innner de fenotipos totales entre los dos datase
ncompartidosall <- nrow(phenopresentall) # Número de fenotipos totales compartidos
if (POR=="jaccard"){
scoreshared <- ncompartidosall/nphenounion
}
if (POR=="forbes"){
scoreshared <- forbes(enrichresult$term_id, targetenrich$term_id)
}
if (adjust == "bonferroni") {
set1join <- dplyr::filter(prunionn, prunionn$bonferroni_pvalue.x <= p.value)
}
if (adjust == "fdr") {
set1join <- dplyr::filter(prunionn, prunionn$FDR_adjust.x <= p.value)
}
set1join <- set1join[,c(1:5,7:12,14:17)]
names(set1join) <- c("term_id", "term_name","source", "bonferroni_pvalue_set1", "FDR_adjust_set1", "gene_overlap_set1", "overlap_ratio_set1", "pvalue_set1", "common_genes_set1","bonferroni_pvalue_set2","FDR_adjust_set2", "gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
set1join$gene_overlap_set1[is.na(set1join$gene_overlap_set1)] <- 0
set1join$gene_overlap_set2[is.na(set1join$gene_overlap_set2)] <- 0
set1join[is.na(set1join)] <- 1
if (adjust == "bonferroni") {
tabledif1u <- dplyr::filter(set1join, set1join$bonferroni_pvalue_set2 > p.value)
# The same for genesetcomaper
set2join <- dplyr::filter(prunionn, prunionn$bonferroni_pvalue.y <= p.value)
}
if (adjust == "fdr") {
tabledif1u <- dplyr::filter(set1join, set1join$FDR_adjust_set2 > p.value)
# The same for genesetcomaper
set2join <- dplyr::filter(prunionn, prunionn$FDR_adjust.y <= p.value)
}
set2join <- set2join[,c(1:5,7:12,14:17)]
names(set2join) <- c("term_id", "term_name","source", "bonferroni_pvalue_set1", "FDR_adjust_set1", "gene_overlap_set1", "overlap_ratio_set1", "pvalue_set1", "common_genes_set1","bonferroni_pvalue_set2","FDR_adjust_set2", "gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
set2join$gene_overlap_set1[is.na(set2join$gene_overlap_set1)] <- 0
set2join$gene_overlap_set2[is.na(set2join$gene_overlap_set2)] <- 0
set2join[is.na(set2join)] <- 1
if (adjust == "bonferroni") {
tabledif2u <- dplyr::filter(set2join, set2join$bonferroni_pvalue_set1 > p.value)
}
if (adjust == "fdr") {
tabledif2u <- dplyr::filter(set2join, set2join$FDR_adjust_set1 > p.value)
}
tabledif2u <- tabledif2u[order(tabledif2u$bonferroni_pvalue_set2),]
tabledif1u <- tabledif1u[order(tabledif1u$bonferroni_pvalue_set1),]
# Unir estos dos datasets
setfusion <- rbind(set1join,set2join)
setfusion <- unique(setfusion)
# Comparación de fenotipos enriquecidos en ambos dataset, número de compartidos
relevantesunion <- full_join(relevantes,targetrelevant, by = "term_id")
nrelevantesunion <- as.numeric(nrow(relevantesunion))
relevantpresent <- inner_join(relevantes,targetrelevant, by = "term_id")
ncompartidosr <- nrow(relevantpresent)
if (POR=="jaccard"){
scoreenrich <- ncompartidosr/nrelevantesunion
}
if (POR=="forbes"){
scoreenrich <- forbes(relevantes$term_id, targetrelevant$term_id)
}
# Inner with all phenotypes
targetenrichttr <- targetenrich[,c(1:5,7:10)]
names(targetenrichttr) <- c("term_id", "term_name", "source", "bonferroni_pvalue_set2", "FDR_adjust_set2","gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
# Table with phenotypes
triucecompare <- merge.data.frame(enrichresult,targetenrichttr, by = c("term_id","term_name", "source"))
if (adjust == "bonferroni") {
# Label if the phenotype is relevant in both datasets
triucecompare <- triucecompare %>%
mutate(shared_enrichment = ifelse(bonferroni_pvalue_set2 <= p.value & bonferroni_pvalue <= p.value, "Yes", "No"))
# All shared phenotypes
triucecompare <- arrange(triucecompare, desc(shared_enrichment), bonferroni_pvalue_set2)
}
if (adjust == "fdr") {
# Label if the phenotype is relevant in both datasets
triucecompare <- triucecompare %>%
mutate(shared_enrichment = ifelse(FDR_adjust_set2 <= p.value & FDR_adjust <= p.value, "Yes", "No"))
# All shared phenotypes
triucecompare <- arrange(triucecompare, desc(shared_enrichment), FDR_adjust_set2)
}
targetrelevanttr <- targetrelevant[,c(1:5,7:10)]
names(targetrelevanttr) <- c("term_id", "term_name", "source","bonferroni_pvalue_set2","FDR_adjust_set2","gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
triucefinal <- merge.data.frame(relevantes,targetrelevanttr, by = c("term_id","term_name", "source"))
triucefinal <- triucefinal[order(triucefinal$bonferroni_pvalue_set),]
nspor <- nrow(triucefinal)
# Esto nos da que cuanto más enriquecido en un sitio más enriquecido en el otro sitio
regresiontarget <- stats::lm(gene_overlap_set1 ~ gene_overlap_set2, data = setfusion, na.action=na.exclude)
ortarget <- summary(regresiontarget)
if (organism == "human") {
nonoverlapgenes <- foundgeneoverlap(geneset, genesetcompare)
}
if (organism == "mouse") {
targetgenesm <- as.data.frame(genesetcompare)
names(targetgenesm) <- "mouse_symbol"
targetgenesm$mouse_symbol <- as.character(targetgenesm$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(targetgenesm, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
targetgenesm = onlygenes
referencegenesm <- as.data.frame(geneset)
names(referencegenesm) <- "mouse_symbol"
referencegenesm$mouse_symbol <- as.character(referencegenesm$mouse_symbol)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(referencegenesm, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
referencegenesm = onlygenes
nonoverlapgenes <- foundgeneoverlap(referencegenesm, targetgenesm)
}
genAll <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/geneproteinlist.txt"))
genAll <- as.data.frame(genAll)
names(genAll) <- c("symbol","entrez")
# We will keep only the selected genes
genAll <- dplyr::anti_join(genAll, nonoverlapgenes, by = "entrez")
genAll <- as.vector(genAll$entrez)
set.seed(seed)
#We will keep a number of a random set of genes with the same number of genes as the genes defined.
nulltestnumbernue = nulltestnumber/2
genesRandom <- replicate(nulltestnumbernue, sample(genAll, length(genesetcompare), replace=TRUE))
genesRandomg2 <- replicate(nulltestnumbernue, sample(genAll, length(geneset), replace=TRUE))
# Detecto los corazones y utilizo la mitad
nworkers <- (detectCores()/2)
# ME HE QUEDADO CON PHENOENRICHCOMPARERANDOM
# Paralelizo la opreación de calcular en enriquecimiento para cada set de genes aleatorios
if (linux==T){
cl <- makeForkCluster(nworkers)
clusterEvalQ(cl, library(PhenoExam))
enrichrandomlist <- parApply(cl, genesRandom, 2, function(x) PhenoEnrichCompareRandom(nonoverlapgenes, geneset, x, database,organism="human"))
enrichrandomlistg2 <- parApply(cl, genesRandomg2, 2, function(x) PhenoEnrichCompareRandom(nonoverlapgenes, geneset, x,database,organism="human"))
stopCluster(cl)
}else{ enrichrandomlist <- apply(genesRandom, 2, function(x) PhenoEnrichCompareRandom(nonoverlapgenes, geneset, x,database,organism="human"))
enrichrandomlistg2 <- apply(genesRandomg2, 2, function(x) PhenoEnrichCompareRandom(nonoverlapgenes, geneset,x,database,organism="human"))}
if (adjust == "bonferroni") {
randomenrichrelevant <- lapply(enrichrandomlist, dplyr::filter, bonferroni_pvalue <= p.value)
randomenrichrelevantg2 <- lapply(enrichrandomlistg2, dplyr::filter, bonferroni_pvalue <= p.value)
}
if (adjust == "fdr") {
randomenrichrelevant <- lapply(enrichrandomlist, dplyr::filter, FDR_adjust <= p.value)
randomenrichrelevantg2 <- lapply(enrichrandomlistg2, dplyr::filter, FDR_adjust <= p.value)
}
# Intersección RPOR
phenopresent <- lapply(enrichrandomlist, inner_join, x = enrichresult, by = "term_name")
ncompartidos <- lapply(phenopresent, nrow)
phenopresentg2 <- lapply(enrichrandomlistg2, inner_join, x = targetenrich, by = "term_name")
ncompartidosg2 <- lapply(phenopresentg2, nrow)
# Unión RPOR
phenopresentunion <- lapply(enrichrandomlist, full_join, x = enrichresult, by = "term_name")
ncompartidosunion <- lapply(phenopresentunion, nrow)
phenopresentuniong2 <- lapply(enrichrandomlistg2, full_join, x = targetenrich, by = "term_name")
ncompartidosuniong2 <- lapply(phenopresentuniong2, nrow)
SharedScore <- mapply(function(x,y) (x/y), x = ncompartidos, y = ncompartidosunion, SIMPLIFY = F)
SharedScoreg2 <- mapply(function(x,y) (x/y), x = ncompartidosg2, y = ncompartidosuniong2, SIMPLIFY = F)
if (POR=="jaccard"){
SharedScore <- c(SharedScore, SharedScoreg2)
}
if (POR=="forbes"){
relev1 <- lapply(enrichrandomlist, function(x,y) {
forbes(x$term_id,y)
}, y = enrichresult$term_id)
relev2 <-
lapply(enrichrandomlistg2, function(x,y) {
forbes(x$term_id,y)
}, y = targetenrich$term_id)
SharedScore <- c(relev1, relev2)
}
# Intersección POR
colnames <- c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust","gene_associated_in_db", "gene_overlap", "overlap_ratio", "pvalue")
randomenrichrelevant <- lapply(randomenrichrelevant, setNames, colnames)
relevantpresent <- lapply(randomenrichrelevant, inner_join, x = relevantes, by = "term_name")
ncompartidosr <- lapply(relevantpresent, nrow)
randomenrichrelevantg2 <- lapply(randomenrichrelevantg2, setNames, colnames)
relevantpresentg2 <- lapply(randomenrichrelevantg2, inner_join, x = targetrelevant, by = "term_name")
ncompartidosrg2 <- lapply(relevantpresentg2, nrow)
# Unión POR
relevantpresentunion <- lapply(randomenrichrelevant, full_join, x = relevantes, by = "term_name")
ncompartidosrunion <- lapply(relevantpresentunion, nrow)
relevantpresentuniong2 <- lapply(randomenrichrelevantg2, full_join, x = targetrelevant, by = "term_name")
ncompartidosruniong2 <- lapply(relevantpresentuniong2, nrow)
EnrichScore <- mapply(function(x,y) (x/y), x = ncompartidosr, y = ncompartidosrunion, SIMPLIFY = F)
EnrichScoreg2 <- mapply(function(x,y) (x/y), x = ncompartidosrg2, y = ncompartidosruniong2, SIMPLIFY = F)
if (POR=="jaccard"){
EnrichScore <- c(EnrichScore, EnrichScoreg2)
}
if (POR=="forbes"){
porrelev1 <- lapply(randomenrichrelevant, function(x,y) {
forbes(x$term_id,y)
}, y = relevantes$term_id)
porrelev2 <-
lapply(randomenrichrelevantg2, function(x,y) {
forbes(x$term_id,y)
}, y = targetrelevant$term_id)
EnrichScore <- c(porrelev1, porrelev2)
}
}
# We will keep with the phenoratios from random genes greater than the phenoratios from predictions genes
# EnrichScore
EnrichScoreMayores = which(EnrichScore > scoreenrich)
EnrichScoreMayoresL = as.numeric(length(EnrichScoreMayores))
# SharedScore
SharedScoreMayores = which(SharedScore > scoreshared)
SharedScoreMayoresL = as.numeric(length(SharedScoreMayores))
# We are going to calculate the pvalue
pvalorEnrichScore = (1 + EnrichScoreMayoresL) / (1 + as.numeric(length(EnrichScore)))
pvalorSharedScoreMayores = (1 + SharedScoreMayoresL) / (1 + as.numeric(length(SharedScore)))
# Plot Relaxed Phenotypic Overlap Ratio (RPOR)
ratiosRandomHP2 <- unlist(SharedScore, use.names = FALSE)
pr <- as.data.frame(ratiosRandomHP2)
rpre <- as.data.frame(scoreshared)
pl1 <- ggplot2::ggplot(pr, ggplot2::aes(x= ratiosRandomHP2)) + ggplot2::geom_histogram(binwidth= mean(ratiosRandomHP2)/30, position="identity", alpha=0.7, fill="red") + ggplot2::geom_vline( ggplot2::aes(xintercept= scoreshared),
color="blue", linetype="dashed", size=0.7)+
labs(title="Relaxed Phenotypic Overlap Ratio (RPOR) from random genes vs geneset2",x= "Relaxed Phenotypic Overlap Ratio (RPOR)", y = "Frecuency in random gene sets")
a <- ggplot2::ggplot_build(pl1)$layout$panel_scales_y[[1]]$range$range
plotshared <- pl1 + ggplot2::geom_text( ggplot2::aes(scoreshared-(scoreshared/25), mean(a), label = "Relaxed Phenotypic Overlap Ratio (POR) from", vjust=0,angle=90, family="Times")) + ggplot2::geom_text( ggplot2::aes(scoreshared-(scoreshared/55), mean(a) ,label = "geneset2", vjust=0,angle=90, family="Times"))
# Plot Phenotype Overlap Ratio (POR)
ratiosRandomHP2 <- unlist(EnrichScore, use.names = FALSE)
pr <- as.data.frame(ratiosRandomHP2)
rpre <- as.data.frame(scoreenrich)
pl1 <- ggplot2::ggplot(pr, ggplot2::aes(x= ratiosRandomHP2)) + ggplot2::geom_histogram(binwidth= 0.005, position="identity", alpha=0.7, fill="red") + ggplot2::geom_vline( ggplot2::aes(xintercept= scoreenrich),
color="blue", linetype="dashed", size=0.7)+
labs(title="Phenotype Overlap Ratio (POR)
from random genes vs geneset2",x= "Phenotype Overlap Ratio (POR)
", y = "Frecuency in random gene sets")
b <- ggplot2::ggplot_build(pl1)$layout$panel_scales_y[[1]]$range$range
plotenrichs <- pl1 + ggplot2::geom_text( ggplot2::aes(scoreenrich-(scoreenrich/40), mean(b) ,label = "POR from geneset2",vjust=0,angle=90, family="Times"))
# Interactive plot
if (adjust == "bonferroni") {
pdif <- setfusion %>% dplyr::mutate(Enrich = case_when( setfusion$bonferroni_pvalue_set1 <= 0.05 & setfusion$bonferroni_pvalue_set2 <= 0.05 ~ "SharedEnrich",
setfusion$bonferroni_pvalue_set1 > 0.05 ~ "Relevant in geneset2",
setfusion$bonferroni_pvalue_set2 > 0.05 ~ "Relevant in geneset1" )) %>%
# prepare text for tooltip
dplyr::mutate(text = paste("term_id: ", term_id, "\nterm_name: ", term_name, "\npvalue_set1: ", bonferroni_pvalue_set1, "\npvalue_set2: ", bonferroni_pvalue_set2, "\nGeneassociation_geneset1: ", gene_overlap_set1, "\nGeneassociation_geneset2: ", gene_overlap_set2, sep="")) %>%
# Classic ggplot
ggplot2::ggplot( ggplot2::aes(x=gene_overlap_set2, y=gene_overlap_set1, color = Enrich, text=text)) +
ggplot2::theme(text = element_text(size=8)) +
ggplot2::geom_point(alpha = 0.5) +
#viridis::scale_color_viridis(discrete=TRUE, guide=FALSE) +
#ggplot2::theme(legend.position="none") +
labs(title = "Phenotype relevance association analysis for gene sets", x = "Number of genes associated in gene set2", y = "Number of genes associated in gene set1") +
theme(plot.title = element_text(size = 15),
axis.title.x = element_text(size = 10),
axis.title.y = element_text(size = 10))
}
if (adjust == "fdr") {
pdif <- setfusion %>% dplyr::mutate(Enrich = case_when( setfusion$FDR_adjust_set1 <= 0.05 & setfusion$FDR_adjust_set2 <= 0.05 ~ "SharedEnrich",
setfusion$FDR_adjust_set1 > 0.05 ~ "Relevant in geneset2",
setfusion$FDR_adjust_set2 > 0.05 ~ "Relevant in geneset1" )) %>%
# prepare text for tooltip
dplyr::mutate(text = paste("term_id: ", term_id, "\nterm_name: ", term_name, "\npvalue_set1: ", FDR_adjust_set1, "\npvalue_set2: ", FDR_adjust_set2, "\nGeneassociation_geneset1: ", gene_overlap_set1, "\nGeneassociation_geneset2: ", gene_overlap_set2, sep="")) %>%
# Classic ggplot
ggplot2::ggplot( ggplot2::aes(x=gene_overlap_set2, y=gene_overlap_set1, color = Enrich, text=text)) +
ggplot2::theme(text = element_text(size=8)) +
ggplot2::geom_point(alpha = 0.5) +
#viridis::scale_color_viridis(discrete=TRUE, guide=FALSE) +
#ggplot2::theme(legend.position="none") +
labs(title = "Phenotype relevance association analysis for gene sets", x = "Number of genes associated in gene set2", y = "Number of genes associated in gene set1") +
theme(plot.title = element_text(size = 15),
axis.title.x = element_text(size = 10),
axis.title.y = element_text(size = 10))
}
# turn ggplot interactive with plotly
ppdif <- plotly::ggplotly(pdif, tooltip="text", parse = T) %>%
add_lines(x = ~gene_overlap_set2, y = fitted(regresiontarget), name = "Regression line")
if (pvalorEnrichScore <= 0.05) {
mensaje <- paste0("Gene set2 and gene set1 shared ",ncompartidosall, " phenotypic terms (out of ",nphenounion, " unique phenotypic terms in both), that yields a Relaxed Phenotypic Overlap Ratio (RPOR) of ",round(scoreshared,3), " (p < ",round(pvalorSharedScoreMayores,6),"). They shared ",nspor, " significant phenotypic terms (out of ",nrelevantesunion, " unique significant phenotypic terms in both), that yields a Phenotypic Overlap Ratio (POR) of ",round(scoreenrich,3), " (p < ",round(pvalorEnrichScore,6),"). Phenotype relevance association analysis for gene sets (i.e. whether the shared phenotypes are similar in relevance, i.e. in the number of genes associated with them, within each gene set) results in an adjusted R squared of ",round(ortarget$adj.r.squared,3)," ( p < ",ortarget$coefficients[2,4],") which suggests that an important portion of the common phenotypes are similar in relevance. The p-values were obtained through randomization of ",nulltestnumber, " random gene sets. Gene set 1 and gene set 2 are statistically significantly similar to each other in phenotypic terms.")
}
if (pvalorEnrichScore > 0.05) {
mensaje <- paste0("Gene set2 and gene set1 shared ",ncompartidosall, " phenotypic terms (out of ",nphenounion, " unique phenotypic terms in both), that yields a Relaxed Phenotypic Overlap Ratio (RPOR) of ",round(scoreshared,3), " (p < ",round(pvalorSharedScoreMayores,6),"). They shared ",nspor, " significant phenotypic terms (out of ",nrelevantesunion, " unique significant phenotypic terms in both), that yields a Phenotypic Overlap Ratio (POR) of ",round(scoreenrich,3), " (p < ",round(pvalorEnrichScore,6),"). Phenotype relevance association analysis for gene sets (i.e. whether the shared phenotypes are similar in relevance, i.e. in the number of genes associated with them, within each gene set) results in an adjusted R squared of ",round(ortarget$adj.r.squared,3)," ( p < ",ortarget$coefficients[2,4],") . The p-values were obtained through randomization of ",nulltestnumber, " random gene sets. Gene set 1 and gene set 2 are not statistically significantly similar to each other in phenotypic terms.")
}
newList <- list("phenomessage" = mensaje, "data.enrich" = triucefinal, "datatotal" = triucecompare, "RPOR" = plotshared, "POR" = plotenrichs, "randomvalues" = pr, "lm" = ortarget, "r.squared" = round(ortarget$adj.r.squared,5), "lmp.value" = ortarget$coefficients[2,4], "plotdif" = ppdif, "tabledif1" = tabledif1u, "tabledif2" = tabledif2u)
return(newList)
}
#' @title PhenoEnrichGenes
#' @description This function do a phenotype enrichment analysis and obtain a table and a graph
#' @param genes A vector with human gene symbol.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param plotn An integer with the number of term that you want in the plot. The default value is 30.
#' @param url A logical parameter. T if you want interactive links
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoEnrichGenes = function(genes, database = "HPO", organism = "human", plotn=30, url = F){
if (organism == "human") {
# Make two empty df with names
rbjointri = data.frame(matrix(vector(), 0, 10,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue","common_genes"))),
stringsAsFactors=F)
rbjoinhtml <-rbjointri
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loadder")
# Run the loadder
peaload <- get(getnamedb)(genes,organism = "human", plotnumber = plotn, url = url)
# join data frames generated
rbjoinhtml <- rbind(rbjoinhtml, peaload$htmltabla)
rbjointri <- rbind(rbjointri, peaload$alldata)
# Order by pvalue
rbjoinhtml = rbjoinhtml[order(rbjoinhtml$bonferroni_pvalue),] # htmltabla
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
# To plot
triucu = rbjointri[1:plotn,]
}
}
if (organism == "mouse") {
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = unique(onlygenes)
genes = genes$symbol
# Make two empty df with names
rbjointri = data.frame(matrix(vector(), 0, 10,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue","FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue","common_genes"))),
stringsAsFactors=F)
rbjoinhtml <-rbjointri
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loadder")
# Run the loadder
peaload <- get(getnamedb)(genes,organism = "human", plotnumber = plotn, url = url)
# join data frames generated
rbjoinhtml <- rbind(rbjoinhtml, peaload$htmltabla)
rbjointri <- rbind(rbjointri, peaload$alldata)
# Order by pvalue
rbjoinhtml = rbjoinhtml[order(rbjoinhtml$bonferroni_pvalue),] # htmltabla
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
# To plot
triucu = rbjointri[1:plotn,]
}
}
# Esto es para que en la generación del gráfico no salga el no fenotipo
#triucu <- filter(triucu, triucu$term_id != "CRB:XXX")
#triucu <- filter(triucu, triucu$term_id != "wXXX") no pasa nada si
# To plot
p <- triucu %>%
# prepare text for tooltip
dplyr::mutate(text = paste("ID Number: ", term_id, "\nTerm: ", term_name, "\nGene Ratio: ", overlap_ratio, "\nBonferroni pvalue: ", bonferroni_pvalue, "\nFDR_adjust: ", FDR_adjust, "\nCommon genes: ", common_genes)) %>%
# Classic ggplot
ggplot( ggplot2::aes(x= gene_overlap , y= reorder(paste0("",term_id,": ",term_name,""), -bonferroni_pvalue), fill = bonferroni_pvalue, text=text)) +
theme (text = element_text(size= 8)) +
geom_bar(stat="identity", alpha=0.7) +
scale_fill_gradientn(colours = rainbow(2), trans= "log") +
theme(legend.position="right") +
labs(title = "Phenotype Enrichment Analysis", x = "Number of genes associated with the term", y = "", fill = "p.value") + #
theme(plot.title = element_text(size = 15),
axis.title.x = element_text(size = 10),
axis.title.y = element_text(size = 10),
axis.text = element_text(
angle = 0))
# turn ggplot interactive with plotly layout.title
nuevvogg <- ggplotly(p, tooltip="text") %>%
config(mathjax = "cdn",
toImageButtonOptions = list(
format = "svg",
filename = "myplot",
width = 1000,
height = 600
)
)
newList <- list("alldata" = rbjointri, "htmltabla" = rbjoinhtml, "graph" = nuevvogg)
return(newList)
}
#' @title PhenoEnrichRandomOpt
#' @description This function makes a plot with enriched phenotype info.
#' @param genes A vector with human gene symbol.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param plotn An integer with the number of term that you want in the plot. The default value is 30.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoEnrichRandomOpt = function(genes, database = "HPO", organism = "human"){
if (organism == "human") {
# Make one empty df with names
rbjointri = data.frame(matrix(vector(), 0, 9,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue"))),
stringsAsFactors=F)
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loadderOpt")
# Run the loadder
peaload <- get(getnamedb)(genes,organism = "human", plotnumber = plotn)
# join data frames generated
rbjointri <- rbind(rbjointri, peaload)
# Order by pvalue
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
}
}
if (organism == "mouse") {
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = unique(onlygenes)
genes = genes$symbol
# Make one empty df with names
rbjointri = data.frame(matrix(vector(), 0, 9,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue"))),
stringsAsFactors=F)
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loadderOpt")
# Run the loadder
peaload <- get(getnamedb)(genes,organism = "human", plotnumber = plotn)
# join data frames generated
rbjointri <- rbind(rbjointri, peaload)
# Order by pvalue
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
}
}
newList <- rbjointri
return(newList)
}
#' @title convertgenesymbolmessage
#' @description This function give a message with the homologues map
#' @param genes A vector with gene symbol.
#' @param organism A string parameter: "human" if you have human genes and "mouse if you have mouse genes
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
convertgenesymbolmessage = function(genes, organism = "human"){
if (organism == "human") {
geneshomo <- as.data.frame(genes)
names(geneshomo) <- "human_symbol"
geneshomo$human_symbol <- as.character(geneshomo$human_symbol)
geneshomo <- unique(geneshomo)
ngeneshomo <- nrow(geneshomo)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygeneshomo <- dplyr::anti_join(geneshomo, homologia, by = "human_symbol")
onlygeneshomo <- onlygeneshomo$human_symbol
nonly <- length(onlygeneshomo)
onlygeneshomo <- paste(onlygeneshomo,collapse=" ")
porcenta <- round(((nonly/ngeneshomo)*100), digits = 2)
if (porcenta == 0) {
mensaje <- paste0("All genes from the input have mice homologues genes")
}
if (porcenta != 0) {
mensaje <- paste0("PhenoExam could not find homologues of the human genes: ", onlygeneshomo," (",nonly,"/",ngeneshomo," genes) ",porcenta,"% of the input genes")
}
return(mensaje)
}
if (organism == "mouse") {
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
genes <- unique(genes)
ngenes <- nrow(genes)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::anti_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$mouse_symbol
nonly <- length(onlygenes)
onlygenes <- paste(onlygenes,collapse=" ")
porcenta <- round(((nonly/ngenes)*100), digits = 2)
mensaje <- paste0("PhenoExam could not find homologues of the mouse genes: ", onlygenes," (",nonly,"/",ngenes," genes) ",porcenta,"% of the input genes")
return(mensaje)
}
}
#' @title entrezmap
#' @description This function give a message with the entrez map
#' @param genes A vector with gene symbol.
#' @param human A logical parameter if you have human gene symbol
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
entrezmap = function(genes, human = T){
if (human) {
mouse = F
# Covert vector to a daframe
genesentrez <- as.data.frame(genes)
names(genesentrez) <- "Symbol"
ngenes <- nrow(genesentrez)
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genesentrez$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genesentrez <- unique(genes.df)
names(genesentrez) <- c("symbol","entrez")
genesentrez$entrez <- as.numeric(as.character(genesentrez$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/geneproteinlist.txt"), header = T)
names(custom) <- c("symbol","entrez")
# We will keep only the selected genes
onlygenesentrez <- dplyr::anti_join(genesentrez, custom, by = "entrez")
# Ahora si tengo los dos y podrÃa mostrar que genes no se mapean o entrar dentro
symyentrez <- unique(onlygenesentrez[,1:2])
genesnoentrez <- symyentrez$symbol
nonly <- as.numeric(length(genesnoentrez))
genesnoentrez <- paste(genesnoentrez,collapse=" ")
porcenta <- round(((nonly/ngenes)*100), digits = 2)
if (porcenta == 0) {
mensaje <- paste0("All genes from the input will be used")
}
if (porcenta != 0) {
mensaje <- paste0("PhenoExam could not use or find these genes: ", genesnoentrez," (",nonly,"/",ngenes," genes) ",porcenta,"% of the input genes")
}
return(mensaje)
}
}
#' @title PhenoEnrichCompare
#' @description This function is used for RandomComparePheno.
#' @param genes A vector with human gene symbol.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param plotn An integer with the number of term that you want in the plot. The default value is 30.
#' @param url A logical parameter. T if you want interactive links
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
# Función para obtener el enriquecimiento de términos HPO Y CRB
PhenoEnrichCompare = function(geneset, genesetcompare, database = "HPO", organism = "human", plotn=30, plotnumber = 1:plotn, url = F){
if (organism == "human") {
# Find gene overlap
onlygenes <- foundgeneoverlap(geneset, genesetcompare)
# Make two empty df with names
rbjointri = data.frame(matrix(vector(), 0, 10,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue","common_genes"))),
stringsAsFactors=F)
rbjoinhtml <-rbjointri
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loaddercompare")
# Run the loadder
peaload <- get(getnamedb)(onlygenes, genesetcompare,organism = "human", plotnumber = plotn, url = url)
# join data frames generated
rbjoinhtml <- rbind(rbjoinhtml, peaload$htmltabla)
rbjointri <- rbind(rbjointri, peaload$alldata)
# Order by pvalue
rbjoinhtml = rbjoinhtml[order(rbjoinhtml$bonferroni_pvalue),] # htmltabla
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
# To plot
triucu = rbjointri[1:plotn,]
}
}
if (organism == "mouse") {
genes <- as.data.frame(genesetcompare)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genesetcompare = unique(onlygenes)
genesr <- as.data.frame(geneset)
names(genesr) <- "mouse_symbol"
genesr$mouse_symbol <- as.character(genesr$mouse_symbol)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genesr, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genesr = unique(onlygenes)
# Encontrar el no overlap entre unos y otros
onlygenes <- foundgeneoverlap(genesr, genesetcompare)
# Make two empty df with names
rbjointri = data.frame(matrix(vector(), 0, 10,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue","common_genes"))),
stringsAsFactors=F)
rbjoinhtml <-rbjointri
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loaddercompare")
# Run the loadder
peaload <- get(getnamedb)(onlygenes, genesetcompare,organism = "human", plotnumber = plotn, url = url)
# join data frames generated
rbjoinhtml <- rbind(rbjoinhtml, peaload$htmltabla)
rbjointri <- rbind(rbjointri, peaload$alldata)
# Order by pvalue
rbjoinhtml = rbjoinhtml[order(rbjoinhtml$bonferroni_pvalue),] # htmltabla
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
# To plot
triucu = rbjointri[1:plotn,]
}
}
newList <- list("alldata" = rbjointri, "htmltabla" = rbjoinhtml)
return(newList)
}
#' @title PhenoEnrichGenespran
#' @description This function makes a plot with enriched phenotype info.
#' @param genes A vector with human gene symbol.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param plotn An integer with the number of term that you want in the plot. The default value is 30.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
# Función para obtener el enriquecimiento de términos HPO Y CRB
PhenoEnrichGenespran = function(genes, database = "HPO", organism = "human", plotn=30, plotnumber = 1:plotn){
if (organism == "human") {
# Si solo HPO
if (database == "HPO") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "HPO", organism)
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si solo CRB
if (database == "CRB") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(4,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si solo MGD
if (database == "MGD" | database == "MOUSEDB"){
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
# Make total_size_ratio per total
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db,17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si todas
if (database == "ALL") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(4,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
triucusinmgd = triucusin
triucusin = rbind(triucusinhpo,triucusincrb,triucusinmgd)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
if (database == "CRBMGD") {
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(4,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
triucusinmgd = triucusin
triucusin = rbind(triucusincrb,triucusinmgd)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
if (database == "HPOMGD") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
triucusinmgd = triucusin
triucusin = rbind(triucusinhpo,triucusinmgd)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
if (database == "HPOCRB") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_name")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
triucusin = rbind(triucusinhpo,triucusincrb)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si todas
if (database == "HUMANDB") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_name")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
triucusin = rbind(triucusinhpo,triucusincrb)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
p <- triucu %>%
# prepare text for tooltip
dplyr::mutate(text = paste("ID Number: ", term_id, "\nTerm: ", term_name, "\nGene Ratio: ", overlap_ratio, "\nAdjust pvalue: ", adjust_pvalue, "\nCommon genes: ", common_genes)) %>%
# Classic ggplot
ggplot( ggplot2::aes(x= gene_overlap , y= reorder(paste0("",term_id,": ",term_name,""), -adjust_pvalue), fill = adjust_pvalue, text=text)) +
theme (text = element_text(size= 8)) +
geom_bar(stat="identity", alpha=0.7) +
scale_fill_gradientn(colours = rainbow(2), trans= "log") +
theme(legend.position="right") +
labs(title = "Phenotype Enrichment Analysis", x = "Number of genes associated with the term", y = "", fill = "p.value") + #
theme(plot.title = element_text(size = 15),
axis.title.x = element_text(size = 10),
axis.title.y = element_text(size = 10),
axis.text = element_text(
angle = 0))
# turn ggplot interactive with plotly layout.title
nuevvogg <- ggplotly(p, tooltip="text") %>%
config(mathjax = "cdn",
toImageButtonOptions = list(
format = "svg",
filename = "myplot",
width = 1000,
height = 600
)
)
}
if (organism == "mouse") {
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = onlygenes
# Si solo HPO
if (database == "HPO") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "HPO", organism)
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si solo CRB
if (database == "CRB") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(4,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si solo MGD
if (database == "MGD" | database == "MOUSEDB"){
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
# Make total_size_ratio per total
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db,17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si todas
if (database == "ALL") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(4,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
triucusinmgd = triucusin
triucusin = rbind(triucusinhpo,triucusincrb,triucusinmgd)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
if (database == "CRBMGD") {
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(4,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
triucusinmgd = triucusin
triucusin = rbind(triucusincrb,triucusinmgd)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
if (database == "HPOMGD") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
triucusinmgd = triucusin
triucusin = rbind(triucusinhpo,triucusinmgd)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si todas
# Si todas
if (database == "HUMANDB") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_name")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
triucusin = rbind(triucusinhpo,triucusincrb)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
p <- triucu %>%
# prepare text for tooltip
dplyr::mutate(text = paste("ID Number: ", term_id, "\nTerm: ", term_name, "\nGene Ratio: ", overlap_ratio, "\nAdjust pvalue: ", adjust_pvalue, "\nCommon genes: ", common_genes)) %>%
# Classic ggplot
ggplot( ggplot2::aes(x= gene_overlap , y= reorder(paste0("",term_id,": ",term_name,""), -adjust_pvalue), fill = adjust_pvalue, text=text)) +
theme (text = element_text(size= 8)) +
geom_bar(stat="identity", alpha=0.7) +
scale_fill_gradientn(colours = rainbow(2), trans= "log") +
theme(legend.position="right") +
labs(title = "Phenotype Enrichment Analysis", x = "Number of genes associated with the term", y = "", fill = "p.value") + #
theme(plot.title = element_text(size = 15),
axis.title.x = element_text(size = 10),
axis.title.y = element_text(size = 10),
axis.text = element_text(
angle = 0))
# turn ggplot interactive with plotly layout.title
nuevvogg <- ggplotly(p, tooltip="text") %>%
config(mathjax = "cdn",
toImageButtonOptions = list(
format = "svg",
filename = "myplot",
width = 1000,
height = 600
)
)
}
newList <- list("alldata" = triucusin, "graph" = nuevvogg)
return(newList)
}
#' @title PhenoEnrichCompareRandom
#' @description This function makes a plot with enriched phenotype info.
#' @param genes A vector with human gene symbol.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param plotn An integer with the number of term that you want in the plot. The default value is 30.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoEnrichCompareRandom = function(onlygenes, geneset, genesetcompare, database = "HPO", organism = "human", plotn=30, plotnumber = 1:plotn){
if (organism == "human") {
# Make two empty df with names
rbjointri = data.frame(matrix(vector(), 0, 9,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue"))),
stringsAsFactors=F)
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loaddercompareOpt")
# Run the loadder
peaload <- get(getnamedb)(onlygenes, geneset, genesetcompare,organism = "human", plotnumber = plotn)
# join data frames generated
rbjointri <- rbind(rbjointri, peaload)
# Order by pvalue
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
# To plot
triucu = rbjointri[1:plotn,]
}
}
if (organism == "mouse") {
genes <- as.data.frame(genesetcompare)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genesetcompare = unique(onlygenes)
genesr <- as.data.frame(geneset)
names(genesr) <- "mouse_symbol"
genesr$mouse_symbol <- as.character(genesr$mouse_symbol)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genesr, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genesr = unique(onlygenes)
# Encontrar el no overlap entre unos y otros
onlygenes <- foundgeneoverlap(genesr, genesetcompare)
# Make two empty df with names
rbjointri = data.frame(matrix(vector(), 0, 9,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue"))),
stringsAsFactors=F)
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loaddercompareOpt")
# Run the loadder
peaload <- get(getnamedb)(onlygenes, genesetcompare,organism = "human", plotnumber = plotn)
# join data frames generated
rbjointri <- rbind(rbjointri, peaload)
# Order by pvalue
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
# To plot
triucu = rbjointri[1:plotn,]
}
}
newList <- rbjointri
return(newList)
}
#' @title foundgeneoverlap
#' @description This function give no overlapping genes
#' @param geneset A vector with human gene symbol.
#' @param genesetcompare A vector with human gene symbol.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
# Función para obtener el enriquecimiento de términos HPO Y CRB
foundgeneoverlap = function(geneset, genesetcompare){
# Ver cuantos genes no hacen overlap para sacarlos del cálculo del target
# Covert vector to a daframe
referencegenesc <- as.data.frame(geneset)
names(referencegenesc) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(referencegenesc$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
referencegenesc <- unique(genes.df)
names(referencegenesc) <- c("symbol","entrez")
referencegenesc$entrez <- as.numeric(as.character(referencegenesc$entrez))
targetgenesc <- as.data.frame(genesetcompare)
names(targetgenesc) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(targetgenesc$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
targetgenesc <- unique(genes.df)
names(targetgenesc) <- c("symbol","entrez")
targetgenesc$entrez <- as.numeric(as.character(targetgenesc$entrez))
# We will keep only the selected genes
onlygenes <- dplyr::anti_join(referencegenesc, targetgenesc, by = "entrez")
return(onlygenes)
}
#
#' @title PhenoRelation
#' @description This function give no overlapping genes
#' @param genes A vector with human gene symbol.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoRelation = function(enrichresult, genes, p.value = 0.05){
# Tomar los fenotipos estadisticamente significativos
relevantes <- dplyr::filter(enrichresult, enrichresult$adjust_pvalue <= p.value)
p <- relevantes %>% dplyr::mutate(db = case_when(stringr::str_detect(relevantes$term_id, "HP") ~ "HPO",
stringr::str_detect(relevantes$term_id, "MP") ~ "MPO",
stringr::str_detect(relevantes$term_id, "CXXX") ~ "DIS",
stringr::str_detect(relevantes$term_id, "C\\d+") ~ "DIS",
stringr::str_detect(relevantes$term_id, "CRB") ~ "CRB"))
hpo <- dplyr::filter( p, p$db == "HPO")
hponr <- nrow(hpo)
nhpo <-round(hponr * 0.1)
hpo <- hpo[c(1:nhpo),]
dis <- dplyr::filter( p, p$db == "DIS")
disnr <- nrow(dis)
ndis <-round(disnr * 0.2)
dis <- dis[c(1:ndis),]
hpodis <- rbind(hpo,dis)
hpodisnomb <- hpodis$term_id
# Contar los fenotipos significativos
numerpheno <- nrow(relevantes) # relevantes
# Obtener el término de todos ellos
nombresmatriz <- relevantes$term_id
nombresfenotipos <- as.data.frame(nombresmatriz)
names(nombresfenotipos) <- "term_id"
# Crear una matriz con el número de fenotipos relevantes
matpheno = matrix(data=NA, nrow= numerpheno, ncol=numerpheno)
rownames(matpheno) <- nombresmatriz
colnames(matpheno) <- nombresmatriz
# Aqui ya empieza a depender de las bases de datos que estén marcadas, creo que esto puede ser siempre todas
allquerygenes <- QueryGenes(genes, database = "ALL")
# Mis genes
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
listadedf = list()
for (i in 1:numerpheno) {
# Leo el fenotipo
phenotipo <- nombresmatriz[i]
# Me quedo con todos los genes y fenotipos relacionados con esos de mis genes
genesdeesefenotipo <- dplyr::filter(allquerygenes, allquerygenes$term_id == phenotipo)
genesdeesefenotipo <- genesdeesefenotipo$entrez
# De alguna forma parece que pierdo un gen del total?
# Con estos genes ahora tengo que calcular el PhenoGeneNumber
numerosdefenotiposhpo <- PhenoGeneNumberOpt(genesdeesefenotipo, database = "HPO")
numerosdefenotiposmdg <- PhenoGeneNumberOpt(genesdeesefenotipo, database = "MGD")
numerosdefenotiposdis <- PhenoGeneNumberOpt(genesdeesefenotipo, database = "DIS")
numerosdefenotiposcrb <- PhenoGeneNumberOpt(genesdeesefenotipo, database = "CRB")
# Junto los cálculos
numerosdefenotipos <- rbind(numerosdefenotiposhpo,numerosdefenotiposdis,numerosdefenotiposmdg,numerosdefenotiposcrb)
names(nombresfenotipos) <- "term_id"
numerorelevantes <- inner_join(numerosdefenotipos,nombresfenotipos, by = "term_id")
nopresentes <- anti_join(nombresfenotipos,numerorelevantes, by = "term_id")
if (nrow(nopresentes)>0) {
nopresentes$gen_associated_number <- 0
nopresentes$term_id <- as.character(nopresentes$term_id)
nopresentes$gen_associated_number <- as.integer(nopresentes$gen_associated_number)
colnames(nopresentes) <- c("term_id","gene_overlap")
numerorelevantes = numerorelevantes[,c(1,3)]
resultadofenotipo <- rbind.data.frame(numerorelevantes,nopresentes)
names(resultadofenotipo) <- c("term_id", paste0("",phenotipo,"") )
}else{
resultadofenotipo =numerorelevantes[,c(1,3)]
names(resultadofenotipo) <- c("term_id", paste0("",phenotipo,"") )
}
listadedf[[i]] <-resultadofenotipo
}
# Esto se lo come rapido
a <- Reduce(function(...) merge(..., by = "term_id", all=TRUE), listadedf)
# Hacer que el data frame sea una matriz (Guardarla para enseñar porque tarda)
mat1 = data.matrix(a)
rownames(mat1) <-a$term_id
mat1<-mat1[,-1]
# Select mp overexpressed
# select crb overexpressed
matnew <-mat1[,colnames(mat1) %in% hpodisnomb]
# El heatmap ya hace un cluster, podemos sacar ese cluster
# COn el cluster tenemos que ver cuando caen cerca términos de diferentes bases de datos y mostrar eso
# Tenemos colnames y rownames
correlacion <- cor(mat1)
matnew <-correlacion[rownames(correlacion) %in% hpodisnomb,colnames(correlacion) %in% hpodisnomb]
flattenCorrMatrix <- function(cormat, pmat) {
ut <- upper.tri(cormat)
data.frame(
row = rownames(cormat)[row(cormat)[ut]],
column = rownames(cormat)[col(cormat)[ut]],
cor =(cormat)[ut],
p = pmat[ut]
)
}
res2<-rcorr(mat1)
fmatda <- flattenCorrMatrix(res2$r, res2$P)
heatmapfin <- pheatmap(matnew, cutree_rows = 4)
newList <- list("cormatrix" = correlacion, "colum" = fmatda, "heatmap" = heatmapfin)
return(newList)
}
#' @title PhenoGeneNumberSimuok
#' @description This function counts the number of genes per phenotype term from the vector of genes.
#' @param database A string parameter with different options: "ALL", "HUMANDB", "MOUSEDB", "HPO", "CRB" and "MGD"
#' @param genes A vector with human gene symbol.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoGeneNumberSimuok = function(genes, database= "HPO", organism = "human"){
if (organism == "human") {
if (database == "HPO") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
names(custom) <- c("entrez","symbol", "term_id", "term_name")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "symbol")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "DIS") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"), header = T)
custom <- custom[,c(1,2,5,6)]
names(custom) <- c("entrez","symbol", "term_id", "term_name")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "symbol")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CRB") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
names(custom) <- c("entrez","symbol", "term_name", "term_id")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "symbol")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "MGD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "human_symbol"
genes$HumanSymbol <- as.character(genes$human_symbol)
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info" )
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "human_symbol")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name ) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
}
if (organism == "mouse") {
# genes = musedata[1:50,3]
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = onlygenes
if (database == "HPO") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
names(custom) <- c("entrez","symbol", "term_id", "term_name")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CRB") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
names(custom) <- c("entrez","symbol", "term_name", "term_id")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "MGD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "human_symbol"
genes$HumanSymbol <- as.character(genes$human_symbol)
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$human_symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info" )
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name ) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
}
return(numportergenes)
}
#' @title PhenoEnrichGenespransimu
#' @description This function analyses the similarity of phenotypic terms between two groups of genes
#' @param genes A vector with gene symbol, fixed genes or panel genes.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param database A string parameter with different options: "ALL", "HUMANDB", "MOUSEDB", "HPO", "CRB" and "MGD"
#' @param p.value Cut of p value as they are relevant terms
#' @param nulltestnumber A number of random genes subset.
#' @param seed A seed.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoEnrichGenespransimu = function(genes, database = "HPO", organism = "human"){
if (organism == "human") {
# Si solo HPO
if (database == "HPO") {
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, nrow(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
}
# Si solo CRB
if (database == "CRB") {
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, nrow(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
}
# Si solo MGD
if (database == "MGD" | database == "MOUSEDB"){
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, nrow(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
}
if (database == "DIS") {
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes, database = "DIS", organism)
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19722, nrow(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19722, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
}
# Si todas
if (database == "ALL") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, nrow(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, nrow(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, nrow(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
triucusinmgd = triucusin
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes, database = "DIS", organism)
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19722, nrow(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19722, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
triucusindis = triucusin
triucusin = rbind(triucusinhpo,triucusincrb,triucusinmgd,triucusindis)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
}
}
return(triucusin)
}
#' @title HPOloadder
#' @export
HPOloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# Number of genes for HPO
ngeneshpo = 19248
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "HPO", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, ngeneshpo - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, ngeneshpo - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "HPO", organism )
if (url ) {
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Get the url for gene
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique names
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
#listadegenesqg[,c(3,6)]
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url ) {
htmlttriucu$term_id <- paste0("<a href='https://hpo.jax.org/app/browse/term/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
# Return list
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title MGDloadder
#' @export
MGDloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
ngenesmgd = 17895
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "MGD", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, ngenesmgd - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, ngenesmgd - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "MGD", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='http://www.informatics.jax.org/vocab/mp_ontology/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CRBloadder
#' @export
CRBloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "CRB", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19274 - genes_associated_in_db , length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "CRB", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(4,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://crisprbrain.org/simple-screen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CLINGENloadder
#' @export
CLINGENloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingendatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "CLINGEN", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "CLINGEN", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title GENOMICS_ENGLANDloadder
#' @export
GENOMICS_ENGLANDloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "GENOMICS_ENGLAND", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19230 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19230 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "GENOMICS_ENGLAND", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CTDloadder
#' @export
CTDloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humandatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "CTD", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19636 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19636 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "CTD", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title ORPHANETloadder
#' @export
ORPHANETloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "ORPHANET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "ORPHANET", organism )
# Get the url for gene
if (url) {
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title PSYGENETloadder
#' @export
PSYGENETloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "PSYGENET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "PSYGENET", organism )
# Get the url for gene
if (url) {
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CGIloadder
#' @export
CGIloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "CGI", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "CGI", organism )
# Get the url for gene
if (url) {
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title UNIPROTloadder
#' @export
UNIPROTloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "UNIPROT", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19204 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19204 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "UNIPROT", organism )
# Get the url for gene
if (url) {
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title HPOloadderOpt
#' @export
HPOloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "HPO", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19248 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title MGDloadderOpt
#' @export
MGDloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "MGD", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 17895 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title CRBloadderOpt
#' @export
CRBloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "CRB", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19274 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title CLINGENloadderOpt
#' @export
CLINGENloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingendatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "CLINGEN", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title GENOMICS_ENGLANDloadderOpt
#' @export
GENOMICS_ENGLANDloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "GENOMICS_ENGLAND", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19230 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19230 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title CTDloadderOpt
#' @export
CTDloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humandatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "CTD", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19636 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19636 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title ORPHANETloadderOpt
#' @export
ORPHANETloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "ORPHANET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title PSYGENETloadderOpt
#' @export
PSYGENETloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "PSYGENET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title UNIPROTloadderOpt
#' @export
UNIPROTloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "UNIPROT", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19204 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19204 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title CGIloadderOpt
#' @export
CGIloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "CGI", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title HPOloaddercompare
#' @export
HPOloaddercompare = function(onlygenes, genesetcompare, organism = "human", plotnumber = 30, url = F){
# Read data base
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "HPO"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "HPO", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare, database = "HPO", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique names
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://hpo.jax.org/app/browse/term/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
# Return list
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title MGDloaddercompare
#' @export
MGDloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "MGD"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "MGD",organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$entrez)), length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$entrez)), length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "MGD", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='http://www.informatics.jax.org/vocab/mp_ontology/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CRBloaddercompare
#' @export
CRBloaddercompare = function(onlygenes, genesetcompare, organism = "human", plotnumber = 30, url = F){
# Read data base
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CRB"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare,database = "CRB",organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "CRB", organism )
# Get the url for gene
if (url) {
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(4,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://crisprbrain.org/simple-screen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CLINGENloaddercompare
#' @export
CLINGENloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CLINGEN"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "CLINGEN", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "CLINGEN", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title GENOMICS_ENGLANDloaddercompare
#' @export
GENOMICS_ENGLANDloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
# Read data base
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "GENOMIC_ENGLAND"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "GENOMICS_ENGLAND", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "GENOMICS_ENGLAND", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CTDloaddercompare
#' @export
CTDloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CTD"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "CTD", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "CTD", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title ORPHANETloaddercompare
#' @export
ORPHANETloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "ORPHANET"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "ORPHANET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "ORPHANET", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title PSYGENETloaddercompare
#' @export
PSYGENETloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "PSYGENET"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "PSYGENET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "PSYGENET", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title UNIPROTloaddercompare
#' @export
UNIPROTloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "UNIPROT"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "UNIPROT", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "UNIPROT", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CGIloaddercompare
#' @export
CGIloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CGI"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "CGI", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "CGI", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title HPOloaddercompareOpt
#' @export
HPOloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
# Read data base
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "HPO"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "HPO", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title MGDloaddercompareOpt
#' @export
MGDloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info", "source" )
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "MGD"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "MGD",organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$entrez)), length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$entrez)), length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title CRBloaddercompareOpt
#' @export
CRBloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
# Read data base
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CRB"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare,database = "CRB",organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title CLINGENloaddercompareOpt
#' @export
CLINGENloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CLINGEN"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "CLINGEN", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title GENOMICS_ENGLANDloaddercompareOpt
#' @export
GENOMICS_ENGLANDloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
# Read data base
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "GENOMICS_ENGLAND"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "GENOMICS_ENGLAND", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title CTDloaddercompareOpt
#' @export
CTDloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CTD"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "CTD", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title ORPHANETloaddercompareOpt
#' @export
ORPHANETloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "ORPHANET"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "ORPHANET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title PSYGENETloaddercompareOpt
#' @export
PSYGENETloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "PSYGENET"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "PSYGENET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title UNIPROTloaddercompareOpt
#' @export
UNIPROTloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "UNIPROT"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "UNIPROT", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title CGIloaddercompareOpt
#' @export
CGIloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CGI"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "CGI", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title forbes
#' @export
forbes<-function(x,y,corrected) {
if (missing(corrected)) {
corrected <- T
}
if (is.numeric(x) && is.numeric(y) && min(x) == 0 && min(y) == 0 && length(x) == length(y)) {
a <- length(which((x * y) > 0))
b <- length(which(x > 0)) - a
c <- length(which(y > 0)) - a
} else {
a <- length(na.omit(match(x,y)))
b <- length(x) - a
c <- length(y) - a
}
n <- a + b + c
if (corrected == T) {
return(a * (n + sqrt(n))/((a + b) * (a + c) + a * sqrt(n) + (b * c)/2))
} else {
return(a * n/((a + b) * (a + c)))
}
}
alexcis95/PhenoExam documentation built on Jan. 6, 2023, 1:34 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions forbes CGIloaddercompareOpt UNIPROTloaddercompareOpt PSYGENETloaddercompareOpt ORPHANETloaddercompareOpt CTDloaddercompareOpt GENOMICS_ENGLANDloaddercompareOpt CLINGENloaddercompareOpt CRBloaddercompareOpt MGDloaddercompareOpt HPOloaddercompareOpt CGIloaddercompare UNIPROTloaddercompare PSYGENETloaddercompare ORPHANETloaddercompare CTDloaddercompare GENOMICS_ENGLANDloaddercompare CLINGENloaddercompare CRBloaddercompare MGDloaddercompare HPOloaddercompare CGIloadderOpt UNIPROTloadderOpt PSYGENETloadderOpt ORPHANETloadderOpt CTDloadderOpt GENOMICS_ENGLANDloadderOpt CLINGENloadderOpt CRBloadderOpt MGDloadderOpt HPOloadderOpt UNIPROTloadder CGIloadder PSYGENETloadder ORPHANETloadder CTDloadder GENOMICS_ENGLANDloadder CLINGENloadder CRBloadder MGDloadder HPOloadder PhenoEnrichGenespransimu PhenoGeneNumberSimuok PhenoRelation foundgeneoverlap PhenoEnrichCompareRandom PhenoEnrichGenespran PhenoEnrichCompare entrezmap convertgenesymbolmessage PhenoEnrichRandomOpt PhenoEnrichGenes RandomComparePheno PhenoGeneNumberOpt PhenoGeneNumber FindPhenoFromGenes ListGenesPheno findoverlapgenes getdbnames QueryGenes
Documented in CGIloadder CGIloaddercompare CGIloaddercompareOpt CGIloadderOpt CLINGENloadder CLINGENloaddercompare CLINGENloaddercompareOpt CLINGENloadderOpt convertgenesymbolmessage CRBloadder CRBloaddercompare CRBloaddercompareOpt CRBloadderOpt CTDloadder CTDloaddercompare CTDloaddercompareOpt CTDloadderOpt entrezmap findoverlapgenes FindPhenoFromGenes forbes foundgeneoverlap GENOMICS_ENGLANDloadder GENOMICS_ENGLANDloaddercompare GENOMICS_ENGLANDloaddercompareOpt GENOMICS_ENGLANDloadderOpt getdbnames HPOloadder HPOloaddercompare HPOloaddercompareOpt HPOloadderOpt ListGenesPheno MGDloadder MGDloaddercompare MGDloaddercompareOpt MGDloadderOpt ORPHANETloadder ORPHANETloaddercompare ORPHANETloaddercompareOpt ORPHANETloadderOpt PhenoEnrichCompare PhenoEnrichCompareRandom PhenoEnrichGenes PhenoEnrichGenespran PhenoEnrichGenespransimu PhenoEnrichRandomOpt PhenoGeneNumber PhenoGeneNumberOpt PhenoGeneNumberSimuok PhenoRelation PSYGENETloadder PSYGENETloaddercompare PSYGENETloaddercompareOpt PSYGENETloadderOpt QueryGenes RandomComparePheno UNIPROTloadder UNIPROTloaddercompare UNIPROTloaddercompareOpt UNIPROTloadderOpt
#' @title QueryGenes
#' @description It produces a data frame with the phenotypical terms from the genes selected and their entrez ids.
#' @param genes A vector with the genes symbol
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes symbol and "mouse" if you have mouse genes symbol
#' @details This function has developed by Alejandro Cisterna Garcia as part of his PhD mentored by Juan Antonio Botia Blaya
#' @import readr
#' @import data.table
#' @import ggplot2
#' @import stats
#' @import plotly
#' @import ggpubr
#' @import dplyr
#' @import viridis
#' @import clusterProfiler
#' @import parallel
#' @import purrr
#' @import DT
#' @import Hmisc
#' @import pheatmap
#' @export
#'
QueryGenes = function(genes, database = "HPO", organism = "human"){
if (organism == "human") {
if (database == "UNIPROT") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CGI") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "PSYGENET") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "ORPHANET") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CTD") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "GENOMICS_ENGLAND") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CLINGEN") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "DIS") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "HPO") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CRB") {
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(custom, genes, by = "entrez")
}
if (database == "MGD" | database == "MOUSEDB") {
# Read mouse data base
musedata <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"))
names(musedata) <- c("entrez","symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info", "source" )
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his MP
onlyg <- inner_join(musedata, genes, by = "entrez")
onlyg <- onlyg[,c(1,2,5,6)]
}
}
if (organism == "mouse") {
# genes = musedata[1:50,3]
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = unique(onlygenes)
if (database == "UNIPROT") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CGI") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "PSYGENET") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "ORPHANET") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CTD") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "GENOMICS_ENGLAND") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Symbol to entrez
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CLINGEN") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "DIS") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "HPO") {
# Read human data base
hpox <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"))
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(hpox, genes, by = "entrez")
}
if (database == "CRB") {
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his HP
onlyg <- inner_join(custom, genes, by = "entrez")
}
if (database == "MGD" | database == "MOUSEDB") {
# Read mouse data base
musedata <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"))
names(musedata) <- c("entrez","symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info", "source" )
# Convert vector to a dataframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# We will keep only with the genes in the query and his MP
onlyg <- inner_join(musedata, genes, by = "entrez")
onlyg <- onlyg[,c(1,2,5,6)]
}
}
return(onlyg)
}
#' @title getdbnames
#' @description This function show the databases available in PhenoExam package.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
getdbnames = function(){
sources <- c("HPO", "CRB", "MGD", "CLINGEN",
"GENOMICS_ENGLAND", "CTD", "ORPHANET",
"PSYGENET", "CGI", "UNIPROT")
return(sources)
}
#' @title findoverlapgenes
#' @description This function get the genes that are linked to the phenotypes selected.
#' @param phenos A vector with the phenotypes id that you want.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
findoverlapgenes = function(phenos){
# Un archivo base de datos con todas las DB
sdb <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/sdb.csv"), header = T)
# Make an empty df with names
rbjointri = data.frame(matrix(vector(), 0, 5,
dimnames=list(c(), c("entrez", "symbol", "term_name", "term_id", "source"))),
stringsAsFactors=F)
aparece <- length(phenos)
# For each database name in a vector run its loadder
for (ph in phenos) {
fenos = dplyr::filter(sdb, sdb$term_id == ph)
# join data frames generated
rbjointri <- rbind(rbjointri, fenos)
}
genesaparecen <- rbjointri %>% dplyr::group_by(symbol,entrez) %>% dplyr::tally() %>%
dplyr::filter(n >= aparece)
genesoverlap <- unique(genesaparecen$symbol)
return(genesoverlap)
}
#' @title ListGenesPheno
#' @description This function make a list with genes selected and his phenotypical terms
#' @param genes A vector with the genes symbol
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes symbol and "mouse" if you have mouse genes symbol
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
ListGenesPheno = function(genes, database = "HPO", organism = "human"){
# We will keep only with the genes in the query and his HP
onlyg <- QueryGenes(genes, database, organism)
# We make a list with genes and his HP
gbygen <- split(onlyg$term_id , onlyg$symbol)
return(gbygen)
}
#' @title FindPhenoFromGenes
#' @description Get the genes selected and only their phenotypical terms defined by the user.
#' @param genes A vector with genes symbol.
#' @param phenoid A vector with id terms to get.
#' @param database A string parameter with different options: "HPO", "DIS", "CRB" and "MGD".
#' @param organism A string parameter: "human" if you have human genes symbol and "mouse" if you have mouse genes symbol
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
FindPhenoFromGenes = function(genes, phenoid, database = "HPO", organism = "human"){
# We will keep only with the genes in the query and his phenotypic terms
onlyg <- QueryGenes(genes,database, organism )
# Convert vector to a dataframe
phenoid <- as.data.frame(phenoid)
names(phenoid) <- "term_id"
# We only keep genes and particular phenotypic term
onlygenesandpehno <- inner_join(onlyg, phenoid, by = "term_id")
return(onlygenesandpehno)
}
#' @title PhenoGeneNumber
#' @description This function counts the number of genes per phenotype term from the vector of genes.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param genes A vector with human gene symbol.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoGeneNumber = function(genes, database= "HPO", organism = "human"){
if (organism == "human") {
if (database == "UNIPROT") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CGI") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "PSYGENET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "ORPHANET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CTD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "GENOMICS_ENGLAND") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CLINGEN") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "DIS") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "HPO") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CRB") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "MGD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "human_symbol"
genes$HumanSymbol <- as.character(genes$human_symbol)
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$human_symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info" , "source" )
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name ) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
}
if (organism == "mouse") {
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = unique(onlygenes)
if (database == "UNIPROT") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CGI") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "PSYGENET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "ORPHANET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CTD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "GENOMICS_ENGLAND") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CLINGEN") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "DIS") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "HPO") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CRB") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "MGD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "human_symbol"
genes$HumanSymbol <- as.character(genes$human_symbol)
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$human_symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info" , "source" )
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name ) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
}
return(numportergenes)
}
#' @title PhenoGeneNumberOpt
#' @description This function counts the number of genes per phenotype term from the vector of genes.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param genes A vector with human gene symbol.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoGeneNumberOpt = function(genes, database= "HPO", organism = "human"){
if (organism == "human") {
if (database == "UNIPROT") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CGI") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "PSYGENET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "ORPHANET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CTD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "GENOMICS_ENGLAND") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CLINGEN") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "DIS") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "HPO") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CRB") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "MGD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info" , "source" )
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name ) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
}
if (organism == "mouse") {
# genes = musedata[1:50,3]
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = unique(onlygenes)
if (database == "UNIPROT") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CGI") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "PSYGENET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "ORPHANET") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CTD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "GENOMICS_ENGLAND") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CLINGEN") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "DIS") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "HPO") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
names(custom) <- c("entrez","symbol", "term_id", "term_name")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CRB") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
names(custom) <- c("entrez","symbol", "term_name", "term_id")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "MGD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "human_symbol"
genes$HumanSymbol <- as.character(genes$human_symbol)
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$human_symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info" , "source" )
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name ) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
}
return(numportergenes)
}
#' @title RandomComparePheno
#' @description This function measures statistically significant phenotype similarities between gene sets and detects significant differential phenotypes for them.
#' @param geneset A vector with genes symbol.
#' @param genesetcompare A vector with gene symbol.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param p.value Cut of p value as they are relevant terms
#' @param nulltestnumber A number of random genes subset.
#' @param seed A seed.
#' @param url A logical parameter. T if you want interactive links
#' @param linux A logical parameter. T if you use linux SO, this is the default option
#' @param adjust A string parameter. You have two options: "bonferroni" is the default option and you can choose "fdr" for False Discovery Rate.
#' @param POR A string parameter. You have two options: Use the Jaccard index "jaccard" is the default option and you can choose the Forbes coefficient "forbes".
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
RandomComparePheno = function(geneset, genesetcompare, organism = "human", database = "HPO", p.value = 0.05, nulltestnumber = 100, seed = 1, conditional_case = F, url = F, linux=T, adjust = "bonferroni", POR = "jaccard"){
if (conditional_case == F) {
# Get the Phenotype Enrichment Analysis from geneset
referencia = PhenoEnrichGenes(genes=geneset, database, organism, url = url)
enrichresult = as.data.frame(referencia$htmltabla)
# Obtain the significative phenotypes terms from geneset
if (adjust == "bonferroni") {
relevantes <- dplyr::filter(enrichresult, enrichresult$bonferroni_pvalue <= p.value)
}
if (adjust == "fdr") {
relevantes <- dplyr::filter(enrichresult, enrichresult$FDR_adjust <= p.value)
}
# Get the Phenotype Enrichment Analysis from genesetcompare
targetenrich = PhenoEnrichGenes(genesetcompare, database, organism, url = url)
targetenrich = as.data.frame(targetenrich$htmltabla)
nphenotargetall <- nrow(targetenrich)
# Obtain the significative phenotypes terms from genesetcompare
if (adjust == "bonferroni") {
targetrelevant <- dplyr::filter(targetenrich, targetenrich$bonferroni_pvalue <= p.value)
}
if (adjust == "fdr") {
targetrelevant <- dplyr::filter(targetenrich, targetenrich$FDR_adjust <= p.value)
}
nphenotarget <- nrow(targetrelevant)
# Obtain overlap
phenounion <- full_join(enrichresult, targetenrich, by = "term_id")
prunionn <- full_join(enrichresult,targetenrich, by = c("term_id","term_name","source"))
nphenounion <- as.numeric(nrow(phenounion))
# Inner phenotypes
phenopresentall <- inner_join(enrichresult, targetenrich, by = "term_id") # Innner de fenotipos totales entre los dos datase
ncompartidosall <- nrow(phenopresentall) # Número de fenotipos totales compartidos
# RPOR interesccion entre union
if (POR=="jaccard"){
scoreshared <- ncompartidosall/nphenounion
}
if (POR=="forbes"){
scoreshared <- forbes(enrichresult$term_id, targetenrich$term_id)
}
# Obtain union sifnificative
if (adjust == "bonferroni") {
set1join <- dplyr::filter(prunionn, prunionn$bonferroni_pvalue.x <= p.value)
}
if (adjust == "fdr") {
set1join <- dplyr::filter(prunionn, prunionn$FDR_adjust.x <= p.value)
}
set1join <- set1join[,c(1:5,7:12,14:17)]
names(set1join) <- c("term_id", "term_name","source", "bonferroni_pvalue_set1", "FDR_adjust_set1", "gene_overlap_set1", "overlap_ratio_set1", "pvalue_set1", "common_genes_set1","bonferroni_pvalue_set2","FDR_adjust_set2", "gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
# Prepare for differentional analysis
set1join$gene_overlap_set1[is.na(set1join$gene_overlap_set1)] <- 0
set1join$gene_overlap_set2[is.na(set1join$gene_overlap_set2)] <- 0
set1join[is.na(set1join)] <- 1
# Obtain differentional phenotypes from geneset
if (adjust == "bonferroni") {
tabledif1u <- dplyr::filter(set1join, set1join$bonferroni_pvalue_set2 > p.value)
# The same for genesetcomaper
set2join <- dplyr::filter(prunionn, prunionn$bonferroni_pvalue.y <= p.value)
}
if (adjust == "fdr") {
tabledif1u <- dplyr::filter(set1join, set1join$FDR_adjust_set2 > p.value)
# The same for genesetcomaper
set2join <- dplyr::filter(prunionn, prunionn$FDR_adjust.y <= p.value)
}
set2join <- set2join[,c(1:5,7:12,14:17)]
names(set2join) <- c("term_id", "term_name","source", "bonferroni_pvalue_set1", "FDR_adjust_set1", "gene_overlap_set1", "overlap_ratio_set1", "pvalue_set1", "common_genes_set1","bonferroni_pvalue_set2","FDR_adjust_set2", "gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
set2join$gene_overlap_set1[is.na(set2join$gene_overlap_set1)] <- 0
set2join$gene_overlap_set2[is.na(set2join$gene_overlap_set2)] <- 0
set2join[is.na(set2join)] <- 1
if (adjust == "bonferroni") {
tabledif2u <- dplyr::filter(set2join, set2join$bonferroni_pvalue_set1 > p.value)
}
if (adjust == "fdr") {
tabledif2u <- dplyr::filter(set2join, set2join$FDR_adjust_set1 > p.value)
}
# Order differentional phenotypes tables
tabledif2u <- tabledif2u[order(tabledif2u$bonferroni_pvalue_set2),]
tabledif1u <- tabledif1u[order(tabledif1u$bonferroni_pvalue_set1),]
# Get fusion df term
setfusion <- rbind(set1join,set2join)
setfusion <- unique(setfusion)
# get and comapre relevant phenotypes terms
relevantesunion <- full_join(relevantes,targetrelevant, by = "term_id")
runionn <- full_join(relevantes,targetrelevant, by = c("term_id","term_name", "source"))
nrelevantesunion <- as.numeric(nrow(relevantesunion))
relevantpresent <- inner_join(relevantes,targetrelevant, by = "term_id")
ncompartidosr <- nrow(relevantpresent)
# Obtain POR
if (POR=="jaccard"){
scoreenrich <- ncompartidosr/nrelevantesunion
}
# FALTA TOCAR ESTE
if (POR=="forbes"){
scoreenrich <- forbes(relevantes$term_id, targetrelevant$term_id)
}
# Inner with all phenotypes
targetenrichttr <- targetenrich[,c(1:5,7:10)]
names(targetenrichttr) <- c("term_id", "term_name", "source", "bonferroni_pvalue_set2", "FDR_adjust_set2","gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
# Table with phenotypes
triucecompare <- merge.data.frame(enrichresult,targetenrichttr, by = c("term_id","term_name", "source"))
if (adjust == "bonferroni") {
# Label if the phenotype is relevant in both datasets
triucecompare <- triucecompare %>%
mutate(shared_enrichment = ifelse(bonferroni_pvalue_set2 <= p.value & bonferroni_pvalue <= p.value, "Yes", "No"))
# All shared phenotypes
triucecompare <- arrange(triucecompare, desc(shared_enrichment), bonferroni_pvalue_set2)
}
if (adjust == "fdr") {
# Label if the phenotype is relevant in both datasets
triucecompare <- triucecompare %>%
mutate(shared_enrichment = ifelse(FDR_adjust_set2 <= p.value & FDR_adjust <= p.value, "Yes", "No"))
# All shared phenotypes
triucecompare <- arrange(triucecompare, desc(shared_enrichment), FDR_adjust_set2)
}
targetrelevanttr <- targetrelevant[,c(1:5,7:10)]
names(targetrelevanttr) <- c("term_id", "term_name", "source","bonferroni_pvalue_set2","FDR_adjust_set2","gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
triucefinal <- merge.data.frame(relevantes,targetrelevanttr, by = c("term_id","term_name", "source"))
triucefinal <- triucefinal[order(triucefinal$bonferroni_pvalue_set),]
nspor <- nrow(triucefinal)
# Regression analysis
regresiontarget <- stats::lm(gene_overlap_set1 ~ gene_overlap_set2, data = setfusion, na.action=na.exclude)
ortarget <- summary(regresiontarget)
# Preparation for simulation analysis
# We need all genes
genAll <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/geneproteinlist.txt"))
genAll <- as.data.frame(genAll)
names(genAll) <- c("symbol","entrez")
# We transform into a vector of entrez numbers
genAll <- as.vector(genAll$entrez)
# set seed
set.seed(seed)
# To split simulation analysis
nulltestnumbernue = nulltestnumber/2
#We will keep a number of a random set of genes with the same number of genes as the genes defined
genesRandom <- replicate(nulltestnumbernue, sample(genAll, length(genesetcompare), replace=TRUE))
genesRandomg2 <- replicate(nulltestnumbernue, sample(genAll, length(geneset), replace=TRUE))
# Detect cores and work with a half of them
nworkers <- (detectCores()/2)
# Parallel simulation analysis
if (linux==T){
cl <- makeForkCluster(nworkers)
clusterEvalQ(cl, library(PhenoExam))
enrichrandomlist <- parApply(cl, genesRandom, 2, function(x) PhenoEnrichRandomOpt(x,database,organism="human"))
enrichrandomlistg2 <- parApply(cl, genesRandomg2, 2, function(x) PhenoEnrichRandomOpt(x,database,organism="human"))
stopCluster(cl)
}else{ enrichrandomlist <- apply(genesRandom, 2, function(x) PhenoEnrichRandomOpt(x,database,organism="human"))
enrichrandomlistg2 <- apply(genesRandomg2, 2, function(x) PhenoEnrichRandomOpt(x,database,organism="human"))}
if (adjust == "bonferroni") {
randomenrichrelevant <- lapply(enrichrandomlist, dplyr::filter, bonferroni_pvalue <= p.value)
randomenrichrelevantg2 <- lapply(enrichrandomlistg2, dplyr::filter, bonferroni_pvalue <= p.value)
}
if (adjust == "fdr") {
randomenrichrelevant <- lapply(enrichrandomlist, dplyr::filter, FDR_adjust <= p.value)
randomenrichrelevantg2 <- lapply(enrichrandomlistg2, dplyr::filter, FDR_adjust <= p.value)
}
# Intersección RPOR
phenopresent <- lapply(enrichrandomlist, inner_join, x = enrichresult, by = "term_name")
ncompartidos <- lapply(phenopresent, nrow)
phenopresentg2 <- lapply(enrichrandomlistg2, inner_join, x = targetenrich, by = "term_name")
ncompartidosg2 <- lapply(phenopresentg2, nrow)
# Unión RPOR
phenopresentunion <- lapply(enrichrandomlist, full_join, x = enrichresult, by = "term_name")
ncompartidosunion <- lapply(phenopresentunion, nrow)
phenopresentuniong2 <- lapply(enrichrandomlistg2, full_join, x = targetenrich, by = "term_name")
ncompartidosuniong2 <- lapply(phenopresentuniong2, nrow)
SharedScore <- mapply(function(x,y) (x/y), x = ncompartidos, y = ncompartidosunion, SIMPLIFY = F)
SharedScoreg2 <- mapply(function(x,y) (x/y), x = ncompartidosg2, y = ncompartidosuniong2, SIMPLIFY = F)
# Aqui tengo que adaptarlo al forbes
if (POR=="jaccard"){
SharedScore <- c(SharedScore, SharedScoreg2)
}
if (POR=="forbes"){
relev1 <- lapply(enrichrandomlist, function(x,y) {
forbes(x$term_id,y)
}, y = enrichresult$term_id)
relev2 <-
lapply(enrichrandomlistg2, function(x,y) {
forbes(x$term_id,y)
}, y = targetenrich$term_id)
SharedScore <- c(relev1, relev2)
}
# Intersección POR
colnames <- c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust","gene_associated_in_db", "gene_overlap", "overlap_ratio", "pvalue")
randomenrichrelevant <- lapply(randomenrichrelevant, setNames, colnames)
relevantpresent <- lapply(randomenrichrelevant, inner_join, x = relevantes, by = "term_name")
ncompartidosr <- lapply(relevantpresent, nrow)
randomenrichrelevantg2 <- lapply(randomenrichrelevantg2, setNames, colnames)
relevantpresentg2 <- lapply(randomenrichrelevantg2, inner_join, x = targetrelevant, by = "term_name")
ncompartidosrg2 <- lapply(relevantpresentg2, nrow)
# Unión POR
relevantpresentunion <- lapply(randomenrichrelevant, full_join, x = relevantes, by = "term_name")
ncompartidosrunion <- lapply(relevantpresentunion, nrow)
relevantpresentuniong2 <- lapply(randomenrichrelevantg2, full_join, x = targetrelevant, by = "term_name")
ncompartidosruniong2 <- lapply(relevantpresentuniong2, nrow)
EnrichScore <- mapply(function(x,y) (x/y), x = ncompartidosr, y = ncompartidosrunion, SIMPLIFY = F)
EnrichScoreg2 <- mapply(function(x,y) (x/y), x = ncompartidosrg2, y = ncompartidosruniong2, SIMPLIFY = F)
if (POR=="jaccard"){
EnrichScore <- c(EnrichScore, EnrichScoreg2)
}
if (POR=="forbes"){
porrelev1 <- lapply(randomenrichrelevant, function(x,y) {
forbes(x$term_id,y)
}, y = relevantes$term_id)
porrelev2 <-
lapply(randomenrichrelevantg2, function(x,y) {
forbes(x$term_id,y)
}, y = targetrelevant$term_id)
EnrichScore <- c(porrelev1, porrelev2)
}
}
if (conditional_case == T) {
# Obtener el análisis de enriquecimiento del dataset de referencia
referencia = PhenoEnrichGenes(genes=geneset, database, organism, url = url)
enrichresult = as.data.frame(referencia$htmltabla)
# Obtain the significative phenotypes terms from geneset
if (adjust == "bonferroni") {
relevantes <- dplyr::filter(enrichresult, enrichresult$bonferroni_pvalue <= p.value)
}
if (adjust == "fdr") {
relevantes <- dplyr::filter(enrichresult, enrichresult$FDR_adjust <= p.value)
}
# Todos los genes de referencia en los que no exista overlap con el dataset comparado deben salir
# de la ecuación
targetenrich = PhenoEnrichCompare(geneset, genesetcompare, database, organism, url = url)
targetenrich = as.data.frame(targetenrich$htmltabla)
nphenotargetall <- nrow(targetenrich)
# names(targetenrich) <- c("term_id", "term_name", "bonferroni_pvalue","gene_associated_in_db", "total_size_ratio", "gene_overlap", "input_size_ratio", "input_total_fold", "overlap_ratio", "pvalue", "common_genes_set2")
# Tomar los fenotipos estadisticamente significativos
if (adjust == "bonferroni") {
targetrelevant <- dplyr::filter(targetenrich, targetenrich$bonferroni_pvalue <= p.value)
}
if (adjust == "fdr") {
targetrelevant <- dplyr::filter(targetenrich, targetenrich$FDR_adjust <= p.value)
}
nphenotarget <- nrow(targetrelevant)
# Pregunta 1 overlap de términos aunque estos no sean relevantes
phenounion <- full_join(enrichresult, targetenrich, by = "term_id")
prunionn <- full_join(enrichresult,targetenrich, by = c("term_id","term_name", "source"))
nphenounion <- as.numeric(nrow(phenounion))
#nphenounion <- unique(phenounion$term_id)
#nphenounion <- length(nphenounion) # Número de fenotipos totales UNICOS DE LA UNION
phenopresentall <- inner_join(enrichresult, targetenrich, by = "term_id") # Innner de fenotipos totales entre los dos datase
ncompartidosall <- nrow(phenopresentall) # Número de fenotipos totales compartidos
if (POR=="jaccard"){
scoreshared <- ncompartidosall/nphenounion
}
if (POR=="forbes"){
scoreshared <- forbes(enrichresult$term_id, targetenrich$term_id)
}
if (adjust == "bonferroni") {
set1join <- dplyr::filter(prunionn, prunionn$bonferroni_pvalue.x <= p.value)
}
if (adjust == "fdr") {
set1join <- dplyr::filter(prunionn, prunionn$FDR_adjust.x <= p.value)
}
set1join <- set1join[,c(1:5,7:12,14:17)]
names(set1join) <- c("term_id", "term_name","source", "bonferroni_pvalue_set1", "FDR_adjust_set1", "gene_overlap_set1", "overlap_ratio_set1", "pvalue_set1", "common_genes_set1","bonferroni_pvalue_set2","FDR_adjust_set2", "gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
set1join$gene_overlap_set1[is.na(set1join$gene_overlap_set1)] <- 0
set1join$gene_overlap_set2[is.na(set1join$gene_overlap_set2)] <- 0
set1join[is.na(set1join)] <- 1
if (adjust == "bonferroni") {
tabledif1u <- dplyr::filter(set1join, set1join$bonferroni_pvalue_set2 > p.value)
# The same for genesetcomaper
set2join <- dplyr::filter(prunionn, prunionn$bonferroni_pvalue.y <= p.value)
}
if (adjust == "fdr") {
tabledif1u <- dplyr::filter(set1join, set1join$FDR_adjust_set2 > p.value)
# The same for genesetcomaper
set2join <- dplyr::filter(prunionn, prunionn$FDR_adjust.y <= p.value)
}
set2join <- set2join[,c(1:5,7:12,14:17)]
names(set2join) <- c("term_id", "term_name","source", "bonferroni_pvalue_set1", "FDR_adjust_set1", "gene_overlap_set1", "overlap_ratio_set1", "pvalue_set1", "common_genes_set1","bonferroni_pvalue_set2","FDR_adjust_set2", "gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
set2join$gene_overlap_set1[is.na(set2join$gene_overlap_set1)] <- 0
set2join$gene_overlap_set2[is.na(set2join$gene_overlap_set2)] <- 0
set2join[is.na(set2join)] <- 1
if (adjust == "bonferroni") {
tabledif2u <- dplyr::filter(set2join, set2join$bonferroni_pvalue_set1 > p.value)
}
if (adjust == "fdr") {
tabledif2u <- dplyr::filter(set2join, set2join$FDR_adjust_set1 > p.value)
}
tabledif2u <- tabledif2u[order(tabledif2u$bonferroni_pvalue_set2),]
tabledif1u <- tabledif1u[order(tabledif1u$bonferroni_pvalue_set1),]
# Unir estos dos datasets
setfusion <- rbind(set1join,set2join)
setfusion <- unique(setfusion)
# Comparación de fenotipos enriquecidos en ambos dataset, número de compartidos
relevantesunion <- full_join(relevantes,targetrelevant, by = "term_id")
nrelevantesunion <- as.numeric(nrow(relevantesunion))
relevantpresent <- inner_join(relevantes,targetrelevant, by = "term_id")
ncompartidosr <- nrow(relevantpresent)
if (POR=="jaccard"){
scoreenrich <- ncompartidosr/nrelevantesunion
}
if (POR=="forbes"){
scoreenrich <- forbes(relevantes$term_id, targetrelevant$term_id)
}
# Inner with all phenotypes
targetenrichttr <- targetenrich[,c(1:5,7:10)]
names(targetenrichttr) <- c("term_id", "term_name", "source", "bonferroni_pvalue_set2", "FDR_adjust_set2","gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
# Table with phenotypes
triucecompare <- merge.data.frame(enrichresult,targetenrichttr, by = c("term_id","term_name", "source"))
if (adjust == "bonferroni") {
# Label if the phenotype is relevant in both datasets
triucecompare <- triucecompare %>%
mutate(shared_enrichment = ifelse(bonferroni_pvalue_set2 <= p.value & bonferroni_pvalue <= p.value, "Yes", "No"))
# All shared phenotypes
triucecompare <- arrange(triucecompare, desc(shared_enrichment), bonferroni_pvalue_set2)
}
if (adjust == "fdr") {
# Label if the phenotype is relevant in both datasets
triucecompare <- triucecompare %>%
mutate(shared_enrichment = ifelse(FDR_adjust_set2 <= p.value & FDR_adjust <= p.value, "Yes", "No"))
# All shared phenotypes
triucecompare <- arrange(triucecompare, desc(shared_enrichment), FDR_adjust_set2)
}
targetrelevanttr <- targetrelevant[,c(1:5,7:10)]
names(targetrelevanttr) <- c("term_id", "term_name", "source","bonferroni_pvalue_set2","FDR_adjust_set2","gene_overlap_set2", "overlap_ratio_set2", "pvalue_set2", "common_genes_set2")
triucefinal <- merge.data.frame(relevantes,targetrelevanttr, by = c("term_id","term_name", "source"))
triucefinal <- triucefinal[order(triucefinal$bonferroni_pvalue_set),]
nspor <- nrow(triucefinal)
# Esto nos da que cuanto más enriquecido en un sitio más enriquecido en el otro sitio
regresiontarget <- stats::lm(gene_overlap_set1 ~ gene_overlap_set2, data = setfusion, na.action=na.exclude)
ortarget <- summary(regresiontarget)
if (organism == "human") {
nonoverlapgenes <- foundgeneoverlap(geneset, genesetcompare)
}
if (organism == "mouse") {
targetgenesm <- as.data.frame(genesetcompare)
names(targetgenesm) <- "mouse_symbol"
targetgenesm$mouse_symbol <- as.character(targetgenesm$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(targetgenesm, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
targetgenesm = onlygenes
referencegenesm <- as.data.frame(geneset)
names(referencegenesm) <- "mouse_symbol"
referencegenesm$mouse_symbol <- as.character(referencegenesm$mouse_symbol)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(referencegenesm, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
referencegenesm = onlygenes
nonoverlapgenes <- foundgeneoverlap(referencegenesm, targetgenesm)
}
genAll <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/geneproteinlist.txt"))
genAll <- as.data.frame(genAll)
names(genAll) <- c("symbol","entrez")
# We will keep only the selected genes
genAll <- dplyr::anti_join(genAll, nonoverlapgenes, by = "entrez")
genAll <- as.vector(genAll$entrez)
set.seed(seed)
#We will keep a number of a random set of genes with the same number of genes as the genes defined.
nulltestnumbernue = nulltestnumber/2
genesRandom <- replicate(nulltestnumbernue, sample(genAll, length(genesetcompare), replace=TRUE))
genesRandomg2 <- replicate(nulltestnumbernue, sample(genAll, length(geneset), replace=TRUE))
# Detecto los corazones y utilizo la mitad
nworkers <- (detectCores()/2)
# ME HE QUEDADO CON PHENOENRICHCOMPARERANDOM
# Paralelizo la opreación de calcular en enriquecimiento para cada set de genes aleatorios
if (linux==T){
cl <- makeForkCluster(nworkers)
clusterEvalQ(cl, library(PhenoExam))
enrichrandomlist <- parApply(cl, genesRandom, 2, function(x) PhenoEnrichCompareRandom(nonoverlapgenes, geneset, x, database,organism="human"))
enrichrandomlistg2 <- parApply(cl, genesRandomg2, 2, function(x) PhenoEnrichCompareRandom(nonoverlapgenes, geneset, x,database,organism="human"))
stopCluster(cl)
}else{ enrichrandomlist <- apply(genesRandom, 2, function(x) PhenoEnrichCompareRandom(nonoverlapgenes, geneset, x,database,organism="human"))
enrichrandomlistg2 <- apply(genesRandomg2, 2, function(x) PhenoEnrichCompareRandom(nonoverlapgenes, geneset,x,database,organism="human"))}
if (adjust == "bonferroni") {
randomenrichrelevant <- lapply(enrichrandomlist, dplyr::filter, bonferroni_pvalue <= p.value)
randomenrichrelevantg2 <- lapply(enrichrandomlistg2, dplyr::filter, bonferroni_pvalue <= p.value)
}
if (adjust == "fdr") {
randomenrichrelevant <- lapply(enrichrandomlist, dplyr::filter, FDR_adjust <= p.value)
randomenrichrelevantg2 <- lapply(enrichrandomlistg2, dplyr::filter, FDR_adjust <= p.value)
}
# Intersección RPOR
phenopresent <- lapply(enrichrandomlist, inner_join, x = enrichresult, by = "term_name")
ncompartidos <- lapply(phenopresent, nrow)
phenopresentg2 <- lapply(enrichrandomlistg2, inner_join, x = targetenrich, by = "term_name")
ncompartidosg2 <- lapply(phenopresentg2, nrow)
# Unión RPOR
phenopresentunion <- lapply(enrichrandomlist, full_join, x = enrichresult, by = "term_name")
ncompartidosunion <- lapply(phenopresentunion, nrow)
phenopresentuniong2 <- lapply(enrichrandomlistg2, full_join, x = targetenrich, by = "term_name")
ncompartidosuniong2 <- lapply(phenopresentuniong2, nrow)
SharedScore <- mapply(function(x,y) (x/y), x = ncompartidos, y = ncompartidosunion, SIMPLIFY = F)
SharedScoreg2 <- mapply(function(x,y) (x/y), x = ncompartidosg2, y = ncompartidosuniong2, SIMPLIFY = F)
if (POR=="jaccard"){
SharedScore <- c(SharedScore, SharedScoreg2)
}
if (POR=="forbes"){
relev1 <- lapply(enrichrandomlist, function(x,y) {
forbes(x$term_id,y)
}, y = enrichresult$term_id)
relev2 <-
lapply(enrichrandomlistg2, function(x,y) {
forbes(x$term_id,y)
}, y = targetenrich$term_id)
SharedScore <- c(relev1, relev2)
}
# Intersección POR
colnames <- c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust","gene_associated_in_db", "gene_overlap", "overlap_ratio", "pvalue")
randomenrichrelevant <- lapply(randomenrichrelevant, setNames, colnames)
relevantpresent <- lapply(randomenrichrelevant, inner_join, x = relevantes, by = "term_name")
ncompartidosr <- lapply(relevantpresent, nrow)
randomenrichrelevantg2 <- lapply(randomenrichrelevantg2, setNames, colnames)
relevantpresentg2 <- lapply(randomenrichrelevantg2, inner_join, x = targetrelevant, by = "term_name")
ncompartidosrg2 <- lapply(relevantpresentg2, nrow)
# Unión POR
relevantpresentunion <- lapply(randomenrichrelevant, full_join, x = relevantes, by = "term_name")
ncompartidosrunion <- lapply(relevantpresentunion, nrow)
relevantpresentuniong2 <- lapply(randomenrichrelevantg2, full_join, x = targetrelevant, by = "term_name")
ncompartidosruniong2 <- lapply(relevantpresentuniong2, nrow)
EnrichScore <- mapply(function(x,y) (x/y), x = ncompartidosr, y = ncompartidosrunion, SIMPLIFY = F)
EnrichScoreg2 <- mapply(function(x,y) (x/y), x = ncompartidosrg2, y = ncompartidosruniong2, SIMPLIFY = F)
if (POR=="jaccard"){
EnrichScore <- c(EnrichScore, EnrichScoreg2)
}
if (POR=="forbes"){
porrelev1 <- lapply(randomenrichrelevant, function(x,y) {
forbes(x$term_id,y)
}, y = relevantes$term_id)
porrelev2 <-
lapply(randomenrichrelevantg2, function(x,y) {
forbes(x$term_id,y)
}, y = targetrelevant$term_id)
EnrichScore <- c(porrelev1, porrelev2)
}
}
# We will keep with the phenoratios from random genes greater than the phenoratios from predictions genes
# EnrichScore
EnrichScoreMayores = which(EnrichScore > scoreenrich)
EnrichScoreMayoresL = as.numeric(length(EnrichScoreMayores))
# SharedScore
SharedScoreMayores = which(SharedScore > scoreshared)
SharedScoreMayoresL = as.numeric(length(SharedScoreMayores))
# We are going to calculate the pvalue
pvalorEnrichScore = (1 + EnrichScoreMayoresL) / (1 + as.numeric(length(EnrichScore)))
pvalorSharedScoreMayores = (1 + SharedScoreMayoresL) / (1 + as.numeric(length(SharedScore)))
# Plot Relaxed Phenotypic Overlap Ratio (RPOR)
ratiosRandomHP2 <- unlist(SharedScore, use.names = FALSE)
pr <- as.data.frame(ratiosRandomHP2)
rpre <- as.data.frame(scoreshared)
pl1 <- ggplot2::ggplot(pr, ggplot2::aes(x= ratiosRandomHP2)) + ggplot2::geom_histogram(binwidth= mean(ratiosRandomHP2)/30, position="identity", alpha=0.7, fill="red") + ggplot2::geom_vline( ggplot2::aes(xintercept= scoreshared),
color="blue", linetype="dashed", size=0.7)+
labs(title="Relaxed Phenotypic Overlap Ratio (RPOR) from random genes vs geneset2",x= "Relaxed Phenotypic Overlap Ratio (RPOR)", y = "Frecuency in random gene sets")
a <- ggplot2::ggplot_build(pl1)$layout$panel_scales_y[[1]]$range$range
plotshared <- pl1 + ggplot2::geom_text( ggplot2::aes(scoreshared-(scoreshared/25), mean(a), label = "Relaxed Phenotypic Overlap Ratio (POR) from", vjust=0,angle=90, family="Times")) + ggplot2::geom_text( ggplot2::aes(scoreshared-(scoreshared/55), mean(a) ,label = "geneset2", vjust=0,angle=90, family="Times"))
# Plot Phenotype Overlap Ratio (POR)
ratiosRandomHP2 <- unlist(EnrichScore, use.names = FALSE)
pr <- as.data.frame(ratiosRandomHP2)
rpre <- as.data.frame(scoreenrich)
pl1 <- ggplot2::ggplot(pr, ggplot2::aes(x= ratiosRandomHP2)) + ggplot2::geom_histogram(binwidth= 0.005, position="identity", alpha=0.7, fill="red") + ggplot2::geom_vline( ggplot2::aes(xintercept= scoreenrich),
color="blue", linetype="dashed", size=0.7)+
labs(title="Phenotype Overlap Ratio (POR)
from random genes vs geneset2",x= "Phenotype Overlap Ratio (POR)
", y = "Frecuency in random gene sets")
b <- ggplot2::ggplot_build(pl1)$layout$panel_scales_y[[1]]$range$range
plotenrichs <- pl1 + ggplot2::geom_text( ggplot2::aes(scoreenrich-(scoreenrich/40), mean(b) ,label = "POR from geneset2",vjust=0,angle=90, family="Times"))
# Interactive plot
if (adjust == "bonferroni") {
pdif <- setfusion %>% dplyr::mutate(Enrich = case_when( setfusion$bonferroni_pvalue_set1 <= 0.05 & setfusion$bonferroni_pvalue_set2 <= 0.05 ~ "SharedEnrich",
setfusion$bonferroni_pvalue_set1 > 0.05 ~ "Relevant in geneset2",
setfusion$bonferroni_pvalue_set2 > 0.05 ~ "Relevant in geneset1" )) %>%
# prepare text for tooltip
dplyr::mutate(text = paste("term_id: ", term_id, "\nterm_name: ", term_name, "\npvalue_set1: ", bonferroni_pvalue_set1, "\npvalue_set2: ", bonferroni_pvalue_set2, "\nGeneassociation_geneset1: ", gene_overlap_set1, "\nGeneassociation_geneset2: ", gene_overlap_set2, sep="")) %>%
# Classic ggplot
ggplot2::ggplot( ggplot2::aes(x=gene_overlap_set2, y=gene_overlap_set1, color = Enrich, text=text)) +
ggplot2::theme(text = element_text(size=8)) +
ggplot2::geom_point(alpha = 0.5) +
#viridis::scale_color_viridis(discrete=TRUE, guide=FALSE) +
#ggplot2::theme(legend.position="none") +
labs(title = "Phenotype relevance association analysis for gene sets", x = "Number of genes associated in gene set2", y = "Number of genes associated in gene set1") +
theme(plot.title = element_text(size = 15),
axis.title.x = element_text(size = 10),
axis.title.y = element_text(size = 10))
}
if (adjust == "fdr") {
pdif <- setfusion %>% dplyr::mutate(Enrich = case_when( setfusion$FDR_adjust_set1 <= 0.05 & setfusion$FDR_adjust_set2 <= 0.05 ~ "SharedEnrich",
setfusion$FDR_adjust_set1 > 0.05 ~ "Relevant in geneset2",
setfusion$FDR_adjust_set2 > 0.05 ~ "Relevant in geneset1" )) %>%
# prepare text for tooltip
dplyr::mutate(text = paste("term_id: ", term_id, "\nterm_name: ", term_name, "\npvalue_set1: ", FDR_adjust_set1, "\npvalue_set2: ", FDR_adjust_set2, "\nGeneassociation_geneset1: ", gene_overlap_set1, "\nGeneassociation_geneset2: ", gene_overlap_set2, sep="")) %>%
# Classic ggplot
ggplot2::ggplot( ggplot2::aes(x=gene_overlap_set2, y=gene_overlap_set1, color = Enrich, text=text)) +
ggplot2::theme(text = element_text(size=8)) +
ggplot2::geom_point(alpha = 0.5) +
#viridis::scale_color_viridis(discrete=TRUE, guide=FALSE) +
#ggplot2::theme(legend.position="none") +
labs(title = "Phenotype relevance association analysis for gene sets", x = "Number of genes associated in gene set2", y = "Number of genes associated in gene set1") +
theme(plot.title = element_text(size = 15),
axis.title.x = element_text(size = 10),
axis.title.y = element_text(size = 10))
}
# turn ggplot interactive with plotly
ppdif <- plotly::ggplotly(pdif, tooltip="text", parse = T) %>%
add_lines(x = ~gene_overlap_set2, y = fitted(regresiontarget), name = "Regression line")
if (pvalorEnrichScore <= 0.05) {
mensaje <- paste0("Gene set2 and gene set1 shared ",ncompartidosall, " phenotypic terms (out of ",nphenounion, " unique phenotypic terms in both), that yields a Relaxed Phenotypic Overlap Ratio (RPOR) of ",round(scoreshared,3), " (p < ",round(pvalorSharedScoreMayores,6),"). They shared ",nspor, " significant phenotypic terms (out of ",nrelevantesunion, " unique significant phenotypic terms in both), that yields a Phenotypic Overlap Ratio (POR) of ",round(scoreenrich,3), " (p < ",round(pvalorEnrichScore,6),"). Phenotype relevance association analysis for gene sets (i.e. whether the shared phenotypes are similar in relevance, i.e. in the number of genes associated with them, within each gene set) results in an adjusted R squared of ",round(ortarget$adj.r.squared,3)," ( p < ",ortarget$coefficients[2,4],") which suggests that an important portion of the common phenotypes are similar in relevance. The p-values were obtained through randomization of ",nulltestnumber, " random gene sets. Gene set 1 and gene set 2 are statistically significantly similar to each other in phenotypic terms.")
}
if (pvalorEnrichScore > 0.05) {
mensaje <- paste0("Gene set2 and gene set1 shared ",ncompartidosall, " phenotypic terms (out of ",nphenounion, " unique phenotypic terms in both), that yields a Relaxed Phenotypic Overlap Ratio (RPOR) of ",round(scoreshared,3), " (p < ",round(pvalorSharedScoreMayores,6),"). They shared ",nspor, " significant phenotypic terms (out of ",nrelevantesunion, " unique significant phenotypic terms in both), that yields a Phenotypic Overlap Ratio (POR) of ",round(scoreenrich,3), " (p < ",round(pvalorEnrichScore,6),"). Phenotype relevance association analysis for gene sets (i.e. whether the shared phenotypes are similar in relevance, i.e. in the number of genes associated with them, within each gene set) results in an adjusted R squared of ",round(ortarget$adj.r.squared,3)," ( p < ",ortarget$coefficients[2,4],") . The p-values were obtained through randomization of ",nulltestnumber, " random gene sets. Gene set 1 and gene set 2 are not statistically significantly similar to each other in phenotypic terms.")
}
newList <- list("phenomessage" = mensaje, "data.enrich" = triucefinal, "datatotal" = triucecompare, "RPOR" = plotshared, "POR" = plotenrichs, "randomvalues" = pr, "lm" = ortarget, "r.squared" = round(ortarget$adj.r.squared,5), "lmp.value" = ortarget$coefficients[2,4], "plotdif" = ppdif, "tabledif1" = tabledif1u, "tabledif2" = tabledif2u)
return(newList)
}
#' @title PhenoEnrichGenes
#' @description This function do a phenotype enrichment analysis and obtain a table and a graph
#' @param genes A vector with human gene symbol.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param plotn An integer with the number of term that you want in the plot. The default value is 30.
#' @param url A logical parameter. T if you want interactive links
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoEnrichGenes = function(genes, database = "HPO", organism = "human", plotn=30, url = F){
if (organism == "human") {
# Make two empty df with names
rbjointri = data.frame(matrix(vector(), 0, 10,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue","common_genes"))),
stringsAsFactors=F)
rbjoinhtml <-rbjointri
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loadder")
# Run the loadder
peaload <- get(getnamedb)(genes,organism = "human", plotnumber = plotn, url = url)
# join data frames generated
rbjoinhtml <- rbind(rbjoinhtml, peaload$htmltabla)
rbjointri <- rbind(rbjointri, peaload$alldata)
# Order by pvalue
rbjoinhtml = rbjoinhtml[order(rbjoinhtml$bonferroni_pvalue),] # htmltabla
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
# To plot
triucu = rbjointri[1:plotn,]
}
}
if (organism == "mouse") {
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = unique(onlygenes)
genes = genes$symbol
# Make two empty df with names
rbjointri = data.frame(matrix(vector(), 0, 10,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue","FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue","common_genes"))),
stringsAsFactors=F)
rbjoinhtml <-rbjointri
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loadder")
# Run the loadder
peaload <- get(getnamedb)(genes,organism = "human", plotnumber = plotn, url = url)
# join data frames generated
rbjoinhtml <- rbind(rbjoinhtml, peaload$htmltabla)
rbjointri <- rbind(rbjointri, peaload$alldata)
# Order by pvalue
rbjoinhtml = rbjoinhtml[order(rbjoinhtml$bonferroni_pvalue),] # htmltabla
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
# To plot
triucu = rbjointri[1:plotn,]
}
}
# Esto es para que en la generación del gráfico no salga el no fenotipo
#triucu <- filter(triucu, triucu$term_id != "CRB:XXX")
#triucu <- filter(triucu, triucu$term_id != "wXXX") no pasa nada si
# To plot
p <- triucu %>%
# prepare text for tooltip
dplyr::mutate(text = paste("ID Number: ", term_id, "\nTerm: ", term_name, "\nGene Ratio: ", overlap_ratio, "\nBonferroni pvalue: ", bonferroni_pvalue, "\nFDR_adjust: ", FDR_adjust, "\nCommon genes: ", common_genes)) %>%
# Classic ggplot
ggplot( ggplot2::aes(x= gene_overlap , y= reorder(paste0("",term_id,": ",term_name,""), -bonferroni_pvalue), fill = bonferroni_pvalue, text=text)) +
theme (text = element_text(size= 8)) +
geom_bar(stat="identity", alpha=0.7) +
scale_fill_gradientn(colours = rainbow(2), trans= "log") +
theme(legend.position="right") +
labs(title = "Phenotype Enrichment Analysis", x = "Number of genes associated with the term", y = "", fill = "p.value") + #
theme(plot.title = element_text(size = 15),
axis.title.x = element_text(size = 10),
axis.title.y = element_text(size = 10),
axis.text = element_text(
angle = 0))
# turn ggplot interactive with plotly layout.title
nuevvogg <- ggplotly(p, tooltip="text") %>%
config(mathjax = "cdn",
toImageButtonOptions = list(
format = "svg",
filename = "myplot",
width = 1000,
height = 600
)
)
newList <- list("alldata" = rbjointri, "htmltabla" = rbjoinhtml, "graph" = nuevvogg)
return(newList)
}
#' @title PhenoEnrichRandomOpt
#' @description This function makes a plot with enriched phenotype info.
#' @param genes A vector with human gene symbol.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param plotn An integer with the number of term that you want in the plot. The default value is 30.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoEnrichRandomOpt = function(genes, database = "HPO", organism = "human"){
if (organism == "human") {
# Make one empty df with names
rbjointri = data.frame(matrix(vector(), 0, 9,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue"))),
stringsAsFactors=F)
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loadderOpt")
# Run the loadder
peaload <- get(getnamedb)(genes,organism = "human", plotnumber = plotn)
# join data frames generated
rbjointri <- rbind(rbjointri, peaload)
# Order by pvalue
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
}
}
if (organism == "mouse") {
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = unique(onlygenes)
genes = genes$symbol
# Make one empty df with names
rbjointri = data.frame(matrix(vector(), 0, 9,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue"))),
stringsAsFactors=F)
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loadderOpt")
# Run the loadder
peaload <- get(getnamedb)(genes,organism = "human", plotnumber = plotn)
# join data frames generated
rbjointri <- rbind(rbjointri, peaload)
# Order by pvalue
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
}
}
newList <- rbjointri
return(newList)
}
#' @title convertgenesymbolmessage
#' @description This function give a message with the homologues map
#' @param genes A vector with gene symbol.
#' @param organism A string parameter: "human" if you have human genes and "mouse if you have mouse genes
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
convertgenesymbolmessage = function(genes, organism = "human"){
if (organism == "human") {
geneshomo <- as.data.frame(genes)
names(geneshomo) <- "human_symbol"
geneshomo$human_symbol <- as.character(geneshomo$human_symbol)
geneshomo <- unique(geneshomo)
ngeneshomo <- nrow(geneshomo)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygeneshomo <- dplyr::anti_join(geneshomo, homologia, by = "human_symbol")
onlygeneshomo <- onlygeneshomo$human_symbol
nonly <- length(onlygeneshomo)
onlygeneshomo <- paste(onlygeneshomo,collapse=" ")
porcenta <- round(((nonly/ngeneshomo)*100), digits = 2)
if (porcenta == 0) {
mensaje <- paste0("All genes from the input have mice homologues genes")
}
if (porcenta != 0) {
mensaje <- paste0("PhenoExam could not find homologues of the human genes: ", onlygeneshomo," (",nonly,"/",ngeneshomo," genes) ",porcenta,"% of the input genes")
}
return(mensaje)
}
if (organism == "mouse") {
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
genes <- unique(genes)
ngenes <- nrow(genes)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::anti_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$mouse_symbol
nonly <- length(onlygenes)
onlygenes <- paste(onlygenes,collapse=" ")
porcenta <- round(((nonly/ngenes)*100), digits = 2)
mensaje <- paste0("PhenoExam could not find homologues of the mouse genes: ", onlygenes," (",nonly,"/",ngenes," genes) ",porcenta,"% of the input genes")
return(mensaje)
}
}
#' @title entrezmap
#' @description This function give a message with the entrez map
#' @param genes A vector with gene symbol.
#' @param human A logical parameter if you have human gene symbol
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
entrezmap = function(genes, human = T){
if (human) {
mouse = F
# Covert vector to a daframe
genesentrez <- as.data.frame(genes)
names(genesentrez) <- "Symbol"
ngenes <- nrow(genesentrez)
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genesentrez$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genesentrez <- unique(genes.df)
names(genesentrez) <- c("symbol","entrez")
genesentrez$entrez <- as.numeric(as.character(genesentrez$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/geneproteinlist.txt"), header = T)
names(custom) <- c("symbol","entrez")
# We will keep only the selected genes
onlygenesentrez <- dplyr::anti_join(genesentrez, custom, by = "entrez")
# Ahora si tengo los dos y podrÃa mostrar que genes no se mapean o entrar dentro
symyentrez <- unique(onlygenesentrez[,1:2])
genesnoentrez <- symyentrez$symbol
nonly <- as.numeric(length(genesnoentrez))
genesnoentrez <- paste(genesnoentrez,collapse=" ")
porcenta <- round(((nonly/ngenes)*100), digits = 2)
if (porcenta == 0) {
mensaje <- paste0("All genes from the input will be used")
}
if (porcenta != 0) {
mensaje <- paste0("PhenoExam could not use or find these genes: ", genesnoentrez," (",nonly,"/",ngenes," genes) ",porcenta,"% of the input genes")
}
return(mensaje)
}
}
#' @title PhenoEnrichCompare
#' @description This function is used for RandomComparePheno.
#' @param genes A vector with human gene symbol.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param plotn An integer with the number of term that you want in the plot. The default value is 30.
#' @param url A logical parameter. T if you want interactive links
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
# Función para obtener el enriquecimiento de términos HPO Y CRB
PhenoEnrichCompare = function(geneset, genesetcompare, database = "HPO", organism = "human", plotn=30, plotnumber = 1:plotn, url = F){
if (organism == "human") {
# Find gene overlap
onlygenes <- foundgeneoverlap(geneset, genesetcompare)
# Make two empty df with names
rbjointri = data.frame(matrix(vector(), 0, 10,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue","common_genes"))),
stringsAsFactors=F)
rbjoinhtml <-rbjointri
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loaddercompare")
# Run the loadder
peaload <- get(getnamedb)(onlygenes, genesetcompare,organism = "human", plotnumber = plotn, url = url)
# join data frames generated
rbjoinhtml <- rbind(rbjoinhtml, peaload$htmltabla)
rbjointri <- rbind(rbjointri, peaload$alldata)
# Order by pvalue
rbjoinhtml = rbjoinhtml[order(rbjoinhtml$bonferroni_pvalue),] # htmltabla
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
# To plot
triucu = rbjointri[1:plotn,]
}
}
if (organism == "mouse") {
genes <- as.data.frame(genesetcompare)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genesetcompare = unique(onlygenes)
genesr <- as.data.frame(geneset)
names(genesr) <- "mouse_symbol"
genesr$mouse_symbol <- as.character(genesr$mouse_symbol)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genesr, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genesr = unique(onlygenes)
# Encontrar el no overlap entre unos y otros
onlygenes <- foundgeneoverlap(genesr, genesetcompare)
# Make two empty df with names
rbjointri = data.frame(matrix(vector(), 0, 10,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue","common_genes"))),
stringsAsFactors=F)
rbjoinhtml <-rbjointri
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loaddercompare")
# Run the loadder
peaload <- get(getnamedb)(onlygenes, genesetcompare,organism = "human", plotnumber = plotn, url = url)
# join data frames generated
rbjoinhtml <- rbind(rbjoinhtml, peaload$htmltabla)
rbjointri <- rbind(rbjointri, peaload$alldata)
# Order by pvalue
rbjoinhtml = rbjoinhtml[order(rbjoinhtml$bonferroni_pvalue),] # htmltabla
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
# To plot
triucu = rbjointri[1:plotn,]
}
}
newList <- list("alldata" = rbjointri, "htmltabla" = rbjoinhtml)
return(newList)
}
#' @title PhenoEnrichGenespran
#' @description This function makes a plot with enriched phenotype info.
#' @param genes A vector with human gene symbol.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param plotn An integer with the number of term that you want in the plot. The default value is 30.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
# Función para obtener el enriquecimiento de términos HPO Y CRB
PhenoEnrichGenespran = function(genes, database = "HPO", organism = "human", plotn=30, plotnumber = 1:plotn){
if (organism == "human") {
# Si solo HPO
if (database == "HPO") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "HPO", organism)
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si solo CRB
if (database == "CRB") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(4,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si solo MGD
if (database == "MGD" | database == "MOUSEDB"){
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
# Make total_size_ratio per total
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db,17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si todas
if (database == "ALL") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(4,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
triucusinmgd = triucusin
triucusin = rbind(triucusinhpo,triucusincrb,triucusinmgd)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
if (database == "CRBMGD") {
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(4,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
triucusinmgd = triucusin
triucusin = rbind(triucusincrb,triucusinmgd)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
if (database == "HPOMGD") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
triucusinmgd = triucusin
triucusin = rbind(triucusinhpo,triucusinmgd)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
if (database == "HPOCRB") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_name")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
triucusin = rbind(triucusinhpo,triucusincrb)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si todas
if (database == "HUMANDB") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_name")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
triucusin = rbind(triucusinhpo,triucusincrb)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
p <- triucu %>%
# prepare text for tooltip
dplyr::mutate(text = paste("ID Number: ", term_id, "\nTerm: ", term_name, "\nGene Ratio: ", overlap_ratio, "\nAdjust pvalue: ", adjust_pvalue, "\nCommon genes: ", common_genes)) %>%
# Classic ggplot
ggplot( ggplot2::aes(x= gene_overlap , y= reorder(paste0("",term_id,": ",term_name,""), -adjust_pvalue), fill = adjust_pvalue, text=text)) +
theme (text = element_text(size= 8)) +
geom_bar(stat="identity", alpha=0.7) +
scale_fill_gradientn(colours = rainbow(2), trans= "log") +
theme(legend.position="right") +
labs(title = "Phenotype Enrichment Analysis", x = "Number of genes associated with the term", y = "", fill = "p.value") + #
theme(plot.title = element_text(size = 15),
axis.title.x = element_text(size = 10),
axis.title.y = element_text(size = 10),
axis.text = element_text(
angle = 0))
# turn ggplot interactive with plotly layout.title
nuevvogg <- ggplotly(p, tooltip="text") %>%
config(mathjax = "cdn",
toImageButtonOptions = list(
format = "svg",
filename = "myplot",
width = 1000,
height = 600
)
)
}
if (organism == "mouse") {
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = onlygenes
# Si solo HPO
if (database == "HPO") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "HPO", organism)
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si solo CRB
if (database == "CRB") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(4,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si solo MGD
if (database == "MGD" | database == "MOUSEDB"){
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
# Make total_size_ratio per total
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db,17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si todas
if (database == "ALL") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(4,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
triucusinmgd = triucusin
triucusin = rbind(triucusinhpo,triucusincrb,triucusinmgd)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
if (database == "CRBMGD") {
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(4,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
triucusinmgd = triucusin
triucusin = rbind(triucusincrb,triucusinmgd)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
if (database == "HPOMGD") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "MGD", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
triucusinmgd = triucusin
triucusin = rbind(triucusinhpo,triucusinmgd)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
# Si todas
# Si todas
if (database == "HUMANDB") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "HPO", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumber(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, length(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
qglist <- QueryGenes(genes,database = "CRB", organism )
urls <- paste0("https://www.genecards.org/cgi-bin/carddisp.pl?gene=",qglist$symbol,"")
qglist$symbol <- paste0("[", qglist$symbol, "](", urls, ")")
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_name")
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
triucusin = rbind(triucusinhpo,triucusincrb)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
# Esto es para decidir el número de términos para plotear
triucu = triucusin[plotnumber,]
}
p <- triucu %>%
# prepare text for tooltip
dplyr::mutate(text = paste("ID Number: ", term_id, "\nTerm: ", term_name, "\nGene Ratio: ", overlap_ratio, "\nAdjust pvalue: ", adjust_pvalue, "\nCommon genes: ", common_genes)) %>%
# Classic ggplot
ggplot( ggplot2::aes(x= gene_overlap , y= reorder(paste0("",term_id,": ",term_name,""), -adjust_pvalue), fill = adjust_pvalue, text=text)) +
theme (text = element_text(size= 8)) +
geom_bar(stat="identity", alpha=0.7) +
scale_fill_gradientn(colours = rainbow(2), trans= "log") +
theme(legend.position="right") +
labs(title = "Phenotype Enrichment Analysis", x = "Number of genes associated with the term", y = "", fill = "p.value") + #
theme(plot.title = element_text(size = 15),
axis.title.x = element_text(size = 10),
axis.title.y = element_text(size = 10),
axis.text = element_text(
angle = 0))
# turn ggplot interactive with plotly layout.title
nuevvogg <- ggplotly(p, tooltip="text") %>%
config(mathjax = "cdn",
toImageButtonOptions = list(
format = "svg",
filename = "myplot",
width = 1000,
height = 600
)
)
}
newList <- list("alldata" = triucusin, "graph" = nuevvogg)
return(newList)
}
#' @title PhenoEnrichCompareRandom
#' @description This function makes a plot with enriched phenotype info.
#' @param genes A vector with human gene symbol.
#' @param database A string parameter with different options. You can define a vector with the databases.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param plotn An integer with the number of term that you want in the plot. The default value is 30.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoEnrichCompareRandom = function(onlygenes, geneset, genesetcompare, database = "HPO", organism = "human", plotn=30, plotnumber = 1:plotn){
if (organism == "human") {
# Make two empty df with names
rbjointri = data.frame(matrix(vector(), 0, 9,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue"))),
stringsAsFactors=F)
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loaddercompareOpt")
# Run the loadder
peaload <- get(getnamedb)(onlygenes, geneset, genesetcompare,organism = "human", plotnumber = plotn)
# join data frames generated
rbjointri <- rbind(rbjointri, peaload)
# Order by pvalue
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
# To plot
triucu = rbjointri[1:plotn,]
}
}
if (organism == "mouse") {
genes <- as.data.frame(genesetcompare)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genesetcompare = unique(onlygenes)
genesr <- as.data.frame(geneset)
names(genesr) <- "mouse_symbol"
genesr$mouse_symbol <- as.character(genesr$mouse_symbol)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genesr, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genesr = unique(onlygenes)
# Encontrar el no overlap entre unos y otros
onlygenes <- foundgeneoverlap(genesr, genesetcompare)
# Make two empty df with names
rbjointri = data.frame(matrix(vector(), 0, 9,
dimnames=list(c(), c("term_id", "term_name", "source", "bonferroni_pvalue", "FDR_adjust",
"genes_associated_in_db","gene_overlap", "overlap_ratio"
,"pvalue"))),
stringsAsFactors=F)
# For each database name in a vector run its loadder
for (db in database) {
# Get the loadder name
getnamedb <- paste0("",db,"loaddercompareOpt")
# Run the loadder
peaload <- get(getnamedb)(onlygenes, genesetcompare,organism = "human", plotnumber = plotn)
# join data frames generated
rbjointri <- rbind(rbjointri, peaload)
# Order by pvalue
rbjointri = rbjointri[order(rbjointri$bonferroni_pvalue),] # alldata
# To plot
triucu = rbjointri[1:plotn,]
}
}
newList <- rbjointri
return(newList)
}
#' @title foundgeneoverlap
#' @description This function give no overlapping genes
#' @param geneset A vector with human gene symbol.
#' @param genesetcompare A vector with human gene symbol.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
# Función para obtener el enriquecimiento de términos HPO Y CRB
foundgeneoverlap = function(geneset, genesetcompare){
# Ver cuantos genes no hacen overlap para sacarlos del cálculo del target
# Covert vector to a daframe
referencegenesc <- as.data.frame(geneset)
names(referencegenesc) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(referencegenesc$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
referencegenesc <- unique(genes.df)
names(referencegenesc) <- c("symbol","entrez")
referencegenesc$entrez <- as.numeric(as.character(referencegenesc$entrez))
targetgenesc <- as.data.frame(genesetcompare)
names(targetgenesc) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(targetgenesc$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
targetgenesc <- unique(genes.df)
names(targetgenesc) <- c("symbol","entrez")
targetgenesc$entrez <- as.numeric(as.character(targetgenesc$entrez))
# We will keep only the selected genes
onlygenes <- dplyr::anti_join(referencegenesc, targetgenesc, by = "entrez")
return(onlygenes)
}
#
#' @title PhenoRelation
#' @description This function give no overlapping genes
#' @param genes A vector with human gene symbol.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoRelation = function(enrichresult, genes, p.value = 0.05){
# Tomar los fenotipos estadisticamente significativos
relevantes <- dplyr::filter(enrichresult, enrichresult$adjust_pvalue <= p.value)
p <- relevantes %>% dplyr::mutate(db = case_when(stringr::str_detect(relevantes$term_id, "HP") ~ "HPO",
stringr::str_detect(relevantes$term_id, "MP") ~ "MPO",
stringr::str_detect(relevantes$term_id, "CXXX") ~ "DIS",
stringr::str_detect(relevantes$term_id, "C\\d+") ~ "DIS",
stringr::str_detect(relevantes$term_id, "CRB") ~ "CRB"))
hpo <- dplyr::filter( p, p$db == "HPO")
hponr <- nrow(hpo)
nhpo <-round(hponr * 0.1)
hpo <- hpo[c(1:nhpo),]
dis <- dplyr::filter( p, p$db == "DIS")
disnr <- nrow(dis)
ndis <-round(disnr * 0.2)
dis <- dis[c(1:ndis),]
hpodis <- rbind(hpo,dis)
hpodisnomb <- hpodis$term_id
# Contar los fenotipos significativos
numerpheno <- nrow(relevantes) # relevantes
# Obtener el término de todos ellos
nombresmatriz <- relevantes$term_id
nombresfenotipos <- as.data.frame(nombresmatriz)
names(nombresfenotipos) <- "term_id"
# Crear una matriz con el número de fenotipos relevantes
matpheno = matrix(data=NA, nrow= numerpheno, ncol=numerpheno)
rownames(matpheno) <- nombresmatriz
colnames(matpheno) <- nombresmatriz
# Aqui ya empieza a depender de las bases de datos que estén marcadas, creo que esto puede ser siempre todas
allquerygenes <- QueryGenes(genes, database = "ALL")
# Mis genes
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
listadedf = list()
for (i in 1:numerpheno) {
# Leo el fenotipo
phenotipo <- nombresmatriz[i]
# Me quedo con todos los genes y fenotipos relacionados con esos de mis genes
genesdeesefenotipo <- dplyr::filter(allquerygenes, allquerygenes$term_id == phenotipo)
genesdeesefenotipo <- genesdeesefenotipo$entrez
# De alguna forma parece que pierdo un gen del total?
# Con estos genes ahora tengo que calcular el PhenoGeneNumber
numerosdefenotiposhpo <- PhenoGeneNumberOpt(genesdeesefenotipo, database = "HPO")
numerosdefenotiposmdg <- PhenoGeneNumberOpt(genesdeesefenotipo, database = "MGD")
numerosdefenotiposdis <- PhenoGeneNumberOpt(genesdeesefenotipo, database = "DIS")
numerosdefenotiposcrb <- PhenoGeneNumberOpt(genesdeesefenotipo, database = "CRB")
# Junto los cálculos
numerosdefenotipos <- rbind(numerosdefenotiposhpo,numerosdefenotiposdis,numerosdefenotiposmdg,numerosdefenotiposcrb)
names(nombresfenotipos) <- "term_id"
numerorelevantes <- inner_join(numerosdefenotipos,nombresfenotipos, by = "term_id")
nopresentes <- anti_join(nombresfenotipos,numerorelevantes, by = "term_id")
if (nrow(nopresentes)>0) {
nopresentes$gen_associated_number <- 0
nopresentes$term_id <- as.character(nopresentes$term_id)
nopresentes$gen_associated_number <- as.integer(nopresentes$gen_associated_number)
colnames(nopresentes) <- c("term_id","gene_overlap")
numerorelevantes = numerorelevantes[,c(1,3)]
resultadofenotipo <- rbind.data.frame(numerorelevantes,nopresentes)
names(resultadofenotipo) <- c("term_id", paste0("",phenotipo,"") )
}else{
resultadofenotipo =numerorelevantes[,c(1,3)]
names(resultadofenotipo) <- c("term_id", paste0("",phenotipo,"") )
}
listadedf[[i]] <-resultadofenotipo
}
# Esto se lo come rapido
a <- Reduce(function(...) merge(..., by = "term_id", all=TRUE), listadedf)
# Hacer que el data frame sea una matriz (Guardarla para enseñar porque tarda)
mat1 = data.matrix(a)
rownames(mat1) <-a$term_id
mat1<-mat1[,-1]
# Select mp overexpressed
# select crb overexpressed
matnew <-mat1[,colnames(mat1) %in% hpodisnomb]
# El heatmap ya hace un cluster, podemos sacar ese cluster
# COn el cluster tenemos que ver cuando caen cerca términos de diferentes bases de datos y mostrar eso
# Tenemos colnames y rownames
correlacion <- cor(mat1)
matnew <-correlacion[rownames(correlacion) %in% hpodisnomb,colnames(correlacion) %in% hpodisnomb]
flattenCorrMatrix <- function(cormat, pmat) {
ut <- upper.tri(cormat)
data.frame(
row = rownames(cormat)[row(cormat)[ut]],
column = rownames(cormat)[col(cormat)[ut]],
cor =(cormat)[ut],
p = pmat[ut]
)
}
res2<-rcorr(mat1)
fmatda <- flattenCorrMatrix(res2$r, res2$P)
heatmapfin <- pheatmap(matnew, cutree_rows = 4)
newList <- list("cormatrix" = correlacion, "colum" = fmatda, "heatmap" = heatmapfin)
return(newList)
}
#' @title PhenoGeneNumberSimuok
#' @description This function counts the number of genes per phenotype term from the vector of genes.
#' @param database A string parameter with different options: "ALL", "HUMANDB", "MOUSEDB", "HPO", "CRB" and "MGD"
#' @param genes A vector with human gene symbol.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoGeneNumberSimuok = function(genes, database= "HPO", organism = "human"){
if (organism == "human") {
if (database == "HPO") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
names(custom) <- c("entrez","symbol", "term_id", "term_name")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "symbol")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "DIS") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disbasededatos.csv"), header = T)
custom <- custom[,c(1,2,5,6)]
names(custom) <- c("entrez","symbol", "term_id", "term_name")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "symbol")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CRB") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "symbol"
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
names(custom) <- c("entrez","symbol", "term_name", "term_id")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "symbol")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "MGD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "human_symbol"
genes$HumanSymbol <- as.character(genes$human_symbol)
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info" )
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "human_symbol")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name ) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
}
if (organism == "mouse") {
# genes = musedata[1:50,3]
genes <- as.data.frame(genes)
names(genes) <- "mouse_symbol"
genes$mouse_symbol <- as.character(genes$mouse_symbol)
# Read database
homologia <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/homology.csv"), header = T)
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(genes, homologia, by = "mouse_symbol")
onlygenes <- onlygenes$human_symbol
onlygenes <- as.data.frame(onlygenes)
names(onlygenes) <- "symbol"
onlygenes$symbol <- as.character(onlygenes$symbol)
genes = onlygenes
if (database == "HPO") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
names(custom) <- c("entrez","symbol", "term_id", "term_name")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "CRB") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "Symbol"
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$Symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
names(custom) <- c("entrez","symbol", "term_name", "term_id")
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
if (database == "MGD") {
# Covert vector to a daframe
genes <- as.data.frame(genes)
names(genes) <- "human_symbol"
genes$HumanSymbol <- as.character(genes$human_symbol)
# Pasar estos genes a entrezid clusterprofiler
genes.df <- bitr(genes$human_symbol, fromType = "SYMBOL",
toType = c("SYMBOL", "ENTREZID"),
OrgDb = "org.Hs.eg.db")
genes <- unique(genes.df$ENTREZID)
genes <- as.data.frame(genes)
names(genes) <- "entrez"
genes$entrez <- as.numeric(as.character(genes$entrez))
# Read database
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info" )
# We will keep only the selected genes
onlygenes <- dplyr::inner_join(custom, genes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- onlygenes %>% dplyr::group_by(term_id, term_name ) %>% dplyr::tally()
numportergenes <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportergenes) <- c("term_id", "term_name", "gene_overlap")
}
}
return(numportergenes)
}
#' @title PhenoEnrichGenespransimu
#' @description This function analyses the similarity of phenotypic terms between two groups of genes
#' @param genes A vector with gene symbol, fixed genes or panel genes.
#' @param organism A string parameter: "human" if you have human genes and "mouse" if you have mouse genes
#' @param database A string parameter with different options: "ALL", "HUMANDB", "MOUSEDB", "HPO", "CRB" and "MGD"
#' @param p.value Cut of p value as they are relevant terms
#' @param nulltestnumber A number of random genes subset.
#' @param seed A seed.
#' @details This function has developed by Alejandro Cisterna GarcÃa as part of his PhD mentored by Juan Antonio BotÃa Blaya
#' @export
PhenoEnrichGenespransimu = function(genes, database = "HPO", organism = "human"){
if (organism == "human") {
# Si solo HPO
if (database == "HPO") {
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, nrow(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
}
# Si solo CRB
if (database == "CRB") {
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, nrow(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
}
# Si solo MGD
if (database == "MGD" | database == "MOUSEDB"){
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, nrow(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
}
if (database == "DIS") {
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes, database = "DIS", organism)
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19722, nrow(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19722, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
}
# Si todas
if (database == "ALL") {
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes,database = "HPO",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248, nrow(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19248, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
# Aqui guardo la fuente de datos de hpo
triucusinhpo = triucusin
# Aqui empieza CRB
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
names(numportertotal) = c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes,database = "CRB",organism)
names(panel) = c( "term_id", "term_name", "gene_overlap")
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274, nrow(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19274, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
# Aqui guardo la fuente de datos de hpo
triucusincrb = triucusin
# Aqui empieza MGD
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
names(numportertotal) <- c("term_id", "term_name", "genes_associated_in_db")
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes, database = "MGD",organism)
numportertotalunir <- numportertotal
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895, nrow(genes), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 17895, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[,c(1:2,6,3:5,7)]
triucusinmgd = triucusin
# Read data base
numportertotal <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/disdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panel <- PhenoGeneNumberSimuok(genes, database = "DIS", organism)
# Make total_size_ratio per total
#numportertotalunir <- numportertotal[,c(1,2,4)]
numportertotalunir <- numportertotal
#panelunir <- panel[,c(1,2,4)]
panelunir <- panel
names(numportertotalunir) = c( "term_id", "term_name", "genes_associated_in_db")
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Esto es para que almenos tenga anotados 10 genes en ese fenotipo
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(adjust_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19722, nrow(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, 19722, nrow(genes), lower.tail = FALSE))
triucusin = triucusin[order(triucusin$adjust_pvalue),]
triucusin = triucusin[,c(1:2,6,3:5,7)]
triucusindis = triucusin
triucusin = rbind(triucusinhpo,triucusincrb,triucusinmgd,triucusindis)
triucusin = triucusin[order(triucusin$adjust_pvalue),]
}
}
return(triucusin)
}
#' @title HPOloadder
#' @export
HPOloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# Number of genes for HPO
ngeneshpo = 19248
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "HPO", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, ngeneshpo - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, ngeneshpo - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "HPO", organism )
if (url ) {
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Get the url for gene
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique names
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
#listadegenesqg[,c(3,6)]
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url ) {
htmlttriucu$term_id <- paste0("<a href='https://hpo.jax.org/app/browse/term/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
# Return list
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title MGDloadder
#' @export
MGDloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
ngenesmgd = 17895
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "MGD", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, ngenesmgd - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, ngenesmgd - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "MGD", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='http://www.informatics.jax.org/vocab/mp_ontology/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CRBloadder
#' @export
CRBloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "CRB", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19274 - genes_associated_in_db , length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "CRB", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(4,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://crisprbrain.org/simple-screen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CLINGENloadder
#' @export
CLINGENloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingendatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "CLINGEN", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "CLINGEN", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title GENOMICS_ENGLANDloadder
#' @export
GENOMICS_ENGLANDloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "GENOMICS_ENGLAND", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19230 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19230 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "GENOMICS_ENGLAND", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CTDloadder
#' @export
CTDloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humandatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "CTD", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19636 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19636 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "CTD", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title ORPHANETloadder
#' @export
ORPHANETloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "ORPHANET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "ORPHANET", organism )
# Get the url for gene
if (url) {
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title PSYGENETloadder
#' @export
PSYGENETloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "PSYGENET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "PSYGENET", organism )
# Get the url for gene
if (url) {
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CGIloadder
#' @export
CGIloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "CGI", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "CGI", organism )
# Get the url for gene
if (url) {
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title UNIPROTloadder
#' @export
UNIPROTloadder = function(genes, organism = "human", plotnumber = 30, url = F){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genes, database = "UNIPROT", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19204 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19204 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genes,database = "UNIPROT", organism )
# Get the url for gene
if (url) {
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title HPOloadderOpt
#' @export
HPOloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpodatatotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "HPO", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19248 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19248 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title MGDloadderOpt
#' @export
MGDloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/totalmponumber.tsv"), header = T, sep = "\t")
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "MGD", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 17895 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 17895 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title CRBloadderOpt
#' @export
CRBloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrtotalnumber.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "CRB", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19274 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19274 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title CLINGENloadderOpt
#' @export
CLINGENloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingendatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "CLINGEN", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title GENOMICS_ENGLANDloadderOpt
#' @export
GENOMICS_ENGLANDloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "GENOMICS_ENGLAND", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19230 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19230 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title CTDloadderOpt
#' @export
CTDloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humandatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "CTD", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19636 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19636 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title ORPHANETloadderOpt
#' @export
ORPHANETloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "ORPHANET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title PSYGENETloadderOpt
#' @export
PSYGENETloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "PSYGENET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19262 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title UNIPROTloadderOpt
#' @export
UNIPROTloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "UNIPROT", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19204 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19204 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title CGIloadderOpt
#' @export
CGIloadderOpt = function(genes, organism = "human", plotnumber = 30){
# Read data base
numportertotalunir <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIdatostotales.csv"), header = T)
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genes, database = "CGI", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE), method = "bonferroni")) %>%
dplyr::mutate(pvalue = phyper(gene_overlap -1, genes_associated_in_db, 19198 - genes_associated_in_db, length(genes), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
}
#' @title HPOloaddercompare
#' @export
HPOloaddercompare = function(onlygenes, genesetcompare, organism = "human", plotnumber = 30, url = F){
# Read data base
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "HPO"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "HPO", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare, database = "HPO", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique names
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://hpo.jax.org/app/browse/term/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
# Return list
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title MGDloaddercompare
#' @export
MGDloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "MGD"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "MGD",organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$entrez)), length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$entrez)), length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "MGD", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,5)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='http://www.informatics.jax.org/vocab/mp_ontology/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CRBloaddercompare
#' @export
CRBloaddercompare = function(onlygenes, genesetcompare, organism = "human", plotnumber = 30, url = F){
# Read data base
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CRB"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare,database = "CRB",organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "CRB", organism )
# Get the url for gene
if (url) {
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(4,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://crisprbrain.org/simple-screen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CLINGENloaddercompare
#' @export
CLINGENloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CLINGEN"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "CLINGEN", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "CLINGEN", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title GENOMICS_ENGLANDloaddercompare
#' @export
GENOMICS_ENGLANDloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
# Read data base
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "GENOMIC_ENGLAND"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "GENOMICS_ENGLAND", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "GENOMICS_ENGLAND", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CTDloaddercompare
#' @export
CTDloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CTD"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "CTD", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "CTD", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title ORPHANETloaddercompare
#' @export
ORPHANETloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "ORPHANET"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "ORPHANET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "ORPHANET", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title PSYGENETloaddercompare
#' @export
PSYGENETloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "PSYGENET"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "PSYGENET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "PSYGENET", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title UNIPROTloaddercompare
#' @export
UNIPROTloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "UNIPROT"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "UNIPROT", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "UNIPROT", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title CGIloaddercompare
#' @export
CGIloaddercompare = function(onlygenes,genesetcompare, organism = "human", plotnumber = 30, url = F){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CGI"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumber(genesetcompare, database = "CGI", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue= p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
# Get each gene per phenotype
qglist <- QueryGenes(genesetcompare,database = "CGI", organism )
if (url) {
# Get the url for gene
qglist$symbol <- paste0("<a href='https://www.genecards.org/cgi-bin/carddisp.pl?gene=", qglist$symbol,"'>", qglist$symbol,"</a>")
}
# Collapse genes per phenotype
listadegenesqg <- qglist %>%
group_by(term_id) %>%
mutate(common_genes = paste0(symbol, collapse = ", "))
# Unique
listadegenesqg <- unique(listadegenesqg[,c(3,6)])
# Join genes and dataframe
triucusin <- inner_join(triucusin,listadegenesqg, by = "term_id")
# Esto es para decidir el número de términos para plotear
triucu = triucusin[1:plotnumber,]
# Make interactive table with phenotype url
htmlttriucu <- triucusin
if (url) {
htmlttriucu$term_id <- paste0("<a href='https://www.ncbi.nlm.nih.gov/medgen/", htmlttriucu$term_id,"'>", htmlttriucu$term_id,"</a>")
}
newList <- list("alldata" = triucusin, "htmltabla" = htmlttriucu)
return(newList)
}
#' @title HPOloaddercompareOpt
#' @export
HPOloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
# Read data base
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/hpobasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "HPO"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "HPO", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title MGDloaddercompareOpt
#' @export
MGDloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/mgibasefinal.tsv"), header = T, sep = "\t")
names(custom) <- c("entrez","human_symbol", "mouse_symbol", "mgi", "term_id", "term_name", "info", "source" )
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "MGD"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "MGD",organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene total_size_ratio col
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$entrez)), length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$entrez)), length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title CRBloaddercompareOpt
#' @export
CRBloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
# Read data base
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/cbrdatabaseok.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CRB"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare,database = "CRB",organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save phenotypes with at least 10 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 10)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title CLINGENloaddercompareOpt
#' @export
CLINGENloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/clingenbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CLINGEN"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "CLINGEN", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title GENOMICS_ENGLANDloaddercompareOpt
#' @export
GENOMICS_ENGLANDloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
# Read data base
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/GENOMICS_ENGLANDbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "GENOMICS_ENGLAND"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "GENOMICS_ENGLAND", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title CTDloaddercompareOpt
#' @export
CTDloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CTD_humanbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CTD"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "CTD", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title ORPHANETloaddercompareOpt
#' @export
ORPHANETloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/ORPHANETbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "ORPHANET"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "ORPHANET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title PSYGENETloaddercompareOpt
#' @export
PSYGENETloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/PSYGENETbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "PSYGENET"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "PSYGENET", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title UNIPROTloaddercompareOpt
#' @export
UNIPROTloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/UNIPROTbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "UNIPROT"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "UNIPROT", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title CGIloaddercompareOpt
#' @export
CGIloaddercompareOpt = function(onlygenes, geneset, genesetcompare, organism = "human", plotnumber = 30){
custom <- fread(paste0(path.package("PhenoExam", quiet = FALSE),
"/CGIbasededatos.csv"), header = T)
custom <- dplyr::anti_join(custom, onlygenes, by = "entrez")
# We will count the number of genes per phenotype
numportertotal <- custom %>% dplyr::group_by(term_id, term_name) %>% dplyr::tally()
numportertotalunir <- dplyr::arrange(numportertotal, dplyr::desc(numportertotal$n))
names(numportertotalunir) <- c("term_id", "term_name", "genes_associated_in_db")
numportertotalunir$source <- "CGI"
# We will count the number of genes per phenotype
panelunir <- PhenoGeneNumberOpt(genesetcompare, database = "CGI", organism)
# We join the dataframes
triucu <- merge.data.frame(numportertotalunir,panelunir, by = c("term_id","term_name"))
# Save diseases terms with at least 5 genes
triucu <- dplyr::filter(triucu, triucu$genes_associated_in_db >= 5)
# To have two DF with different names
triucusin <- triucu
# Add gene overlap_ratio
triucusin <- triucusin %>%
dplyr::mutate(overlap_ratio = paste0("(",gene_overlap,"/",genes_associated_in_db,")"))
# Calculate p.value (genes in each database)
# Calculate p.value
triucusin <- triucusin %>%
dplyr::mutate(bonferroni_pvalue = p.adjust(phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE), method = "bonferroni"))%>%
dplyr::mutate(pvalue = phyper(gene_overlap - 1, genes_associated_in_db, length(unique(custom$symbol)) - genes_associated_in_db, length(genesetcompare), lower.tail = FALSE))
triucusin <- triucusin %>%
dplyr::mutate(FDR_adjust= p.adjust(p = triucusin$pvalue, method = "fdr"))
# Order by pvalue
triucusin = triucusin[order(triucusin$bonferroni_pvalue),]
triucusin = triucusin[,c(1:2,4,7,9,3,5,6,8)]
return(triucusin)
}
#' @title forbes
#' @export
forbes<-function(x,y,corrected) {
if (missing(corrected)) {
corrected <- T
}
if (is.numeric(x) && is.numeric(y) && min(x) == 0 && min(y) == 0 && length(x) == length(y)) {
a <- length(which((x * y) > 0))
b <- length(which(x > 0)) - a
c <- length(which(y > 0)) - a
} else {
a <- length(na.omit(match(x,y)))
b <- length(x) - a
c <- length(y) - a
}
n <- a + b + c
if (corrected == T) {
return(a * (n + sqrt(n))/((a + b) * (a + c) + a * sqrt(n) + (b * c)/2))
} else {
return(a * n/((a + b) * (a + c)))
}
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.