Functional_Annotation.R
In biomartr: Genomic Data Retrieval

## ----echo = FALSE, message = FALSE------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
options(width = 750)
knitr::opts_chunk$set(
  comment = "#>",
  error = FALSE,
  tidy = FALSE)

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # install the biomaRt package		
#  # source("https://bioconductor.org/biocLite.R")		
#  # biocLite("biomaRt")		
#  # load biomaRt		
#  library(biomaRt)		
#  # look at top 10 databases		
#  head(biomaRt::listMarts(host = "https://www.ensembl.org"), 10)		

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  head(biomaRt::listDatasets(biomaRt::useMart("ENSEMBL_MART_ENSEMBL", host = "https://www.ensembl.org")), 10)		

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  head(biomaRt::listAttributes(biomaRt::useDataset(
#                                           dataset = "hsapiens_gene_ensembl", 		
#                                           mart    = useMart("ENSEMBL_MART_ENSEMBL",		
#                                           host    = "https://www.ensembl.org"))), 10)		

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#   head(biomaRt::listFilters(biomaRt::useDataset(dataset = "hsapiens_gene_ensembl", 		
#                                                 mart    = useMart("ENSEMBL_MART_ENSEMBL",		
#                                                 host    = "https://www.ensembl.org"))), 10)		

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#   # 1) select a mart and data set		
#   mart <- biomaRt::useDataset(dataset = "hsapiens_gene_ensembl", 		
#                      mart    = useMart("ENSEMBL_MART_ENSEMBL",		
#                      host    = "https://www.ensembl.org"))		
#   		
#   # 2) run a biomart query using the getBM() function		
#   # and specify the attributes and filter arguments		
#   geneSet <- "GUCA2A"		
#   		
#   resultTable <- biomaRt::getBM(attributes = c("start_position","end_position","description"),		
#                        filters    = "hgnc_symbol", 		
#                        values     = geneSet, 		
#                        mart       = mart)		
#   		
#   resultTable 		

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # load the biomartr package
#  library(biomartr)
#  
#  # list all available databases
#  biomartr::getMarts()

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  head(biomartr::getDatasets(mart = "ENSEMBL_MART_ENSEMBL") , 5)

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  tail(biomartr::getDatasets(mart = "ENSEMBL_MART_ENSEMBL") , 38)

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # show all elements of the data.frame
#  options(tibble.print_max = Inf)
#  # list all available attributes for dataset: hsapiens_gene_ensembl
#  head( biomartr::getAttributes(mart    = "ENSEMBL_MART_ENSEMBL",
#                                dataset = "hsapiens_gene_ensembl"), 10 )

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # show all elements of the data.frame
#  options(tibble.print_max = Inf)
#  # list all available filters for dataset: hsapiens_gene_ensembl
#  head( biomartr::getFilters(mart    = "ENSEMBL_MART_ENSEMBL",
#                             dataset = "hsapiens_gene_ensembl"), 10 )

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # show all elements of the data.frame
#  options(tibble.print_max = Inf)
#  # retrieving all available datasets and biomart connections for
#  # a specific query organism (scientific name)
#  biomartr::organismBM(organism = "Homo sapiens")

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # show all elements of the data.frame
#  options(tibble.print_max = Inf)
#  # return available attributes for "Homo sapiens"
#  head(biomartr::organismAttributes("Homo sapiens"), 20)

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # show all elements of the data.frame
#  options(tibble.print_max = Inf)
#  # search for attribute topic "id"
#  head(biomartr::organismAttributes("Homo sapiens", topic = "id"), 20)

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # show all elements of the data.frame
#  options(tibble.print_max = Inf)
#  # search for attribute topic "homolog"
#  head(biomartr::organismAttributes("Homo sapiens", topic = "homolog"), 20)

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # show all elements of the data.frame
#  options(tibble.print_max = Inf)
#  # search for attribute topic "dn"
#  head(biomartr::organismAttributes("Homo sapiens", topic = "dn"))

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # show all elements of the data.frame
#  options(tibble.print_max = Inf)
#  # search for attribute topic "ds"
#  head(biomartr::organismAttributes("Homo sapiens", topic = "ds"))

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # show all elements of the data.frame
#  options(tibble.print_max = Inf)
#  # return available filters for "Homo sapiens"
#  head(biomartr::organismFilters("Homo sapiens"), 20)

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # show all elements of the data.frame
#  options(tibble.print_max = Inf)
#  # search for filter topic "id"
#  head(biomartr::organismFilters("Homo sapiens", topic = "id"), 20)

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # show all elements of the data.frame
#  options(tibble.print_max = Inf)
#  
#  head(biomartr::organismAttributes("Homo sapiens", topic = "id"))

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # show all elements of the data.frame
#  options(tibble.print_max = Inf)
#  # retrieve the proteome of Homo sapiens from refseq
#  file_path <- biomartr::getProteome( db       = "refseq",
#                                      organism = "Homo sapiens",
#                                      path     = file.path("_ncbi_downloads","proteomes") )
#  
#  Hsapiens_proteome <- biomartr::read_proteome(file_path, format = "fasta")
#  
#  # remove splice variants from id
#  gene_set <- unlist(sapply(strsplit(Hsapiens_proteome@ranges@NAMES[1:5], ".",fixed = TRUE), function(x) x[1]))
#  
#  result_BM <- biomartr::biomart( genes      = gene_set, # genes were retrieved using biomartr::getGenome()
#                                  mart       = "ENSEMBL_MART_ENSEMBL", # marts were selected with biomartr::getMarts()
#                                  dataset    = "hsapiens_gene_ensembl", # datasets were selected with biomartr::getDatasets()
#                                  attributes = c("ensembl_gene_id","ensembl_peptide_id"), # attributes were selected with biomartr::getAttributes()
#                                  filters    = "refseq_peptide") # specify what ID type was stored in the fasta file retrieved with biomartr::getGenome()
#  
#  result_BM

## ----eval=FALSE-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
#  # show all elements of the data.frame
#  options(tibble.print_max = Inf)
#  # search for GO terms of an example Homo sapiens gene
#  GO_tbl <- biomartr::getGO(organism = "Homo sapiens",
#                            genes    = "GUCA2A",
#                            filters  = "hgnc_symbol")
#  
#  GO_tbl