R/Functions_class_featureTable.R
In mjhelf/Mosaic: Metaboseek platform
Defines functions
asMetaboAnalyst
asInclusionlist
tableWriter
tableGrouping
updateFTIndex
updateFTgrouping
updateFeatureTable
.reformatFeatureTable
.getFTFormat
updateDF
constructFeatureTable
Documented in
asInclusionlist
asMetaboAnalyst
constructFeatureTable
.getFTFormat
.reformatFeatureTable
tableGrouping
tableWriter
updateDF
updateFeatureTable
updateFTgrouping
updateFTIndex
#' constructFeatureTable
#'
#'
#' Constructor function for "MseekFT" object from a feature table data.frame.
#' uses and retains the original data frame plus names of columns containing
#' relevant data.
#'
#' NOTE: currently, only the default values for column names are supported.
#'
#' @param df feature table as data.frame, with retention time (rt), mz and
#' (optionally) intensity values.
#' @param mzcol column in df with mz values (columnname), defaults to "mz"
#' @param rtcol column in df with rt values (columnname), defaults to "rt"
#' @param commentcol column in df with comments (columnname), defaults to "comments"
#' @param fragmentcol column in df with fragmentation information (columnname),
#' defaults to "fragments"
#' @param rtFormat Are retention times given in seconds ("sec") or minutes ("min")
#' @param anagrouptable Analysis grouping table: a data.frame with columns
#' "Column" (containing column names from df with intensity values) and "Group"
#' (defining a group for each entry in "Column")
#' @param tablename Name of the table as displayed by Mseek
#' @param editable allow editing of this table in the Mseek app? if FALSE, only
#' comments column can be edited. editable tables are also not paginated.
#' @param processHistory a list of \code{\link[xcms]{processHistory}} objects
#'
#' @return an \code{MseekFT} object containing a feature table and metadata
#'
#' @export
constructFeatureTable <- function(df= data.frame(mz=numeric(3), rt = numeric(3)),# data frame
mzcol= "mz", #
rtcol= "rt", #column in df with mz values (columnname)
commentcol = "comments",
fragmentcol = "fragments",
rtFormat = "sec", # "sec" or "min"
anagrouptable = NULL,
tablename = "Custom Table",
editable = T,
processHistory = list()){ #T: free editing (add rows), but always see all columns in viewer, F: only comments can be edited directly, no adding of columns
#make columns if they don't exist:
if (class(try(df[,mzcol], silent = T))=="try-error"){ df[,mzcol] <- numeric(nrow(df)) }
if (class(try(df[,rtcol], silent = T))=="try-error"){ df[,rtcol] <- numeric(nrow(df)) }
if (class(try(df[,commentcol], silent = T))=="try-error"){ df[,commentcol] <- character(nrow(df)) }
if (class(try(df[,fragmentcol], silent = T))=="try-error"){ df[,fragmentcol] <- character(nrow(df)) }
FT = list()
#save the entire dataframe
FT$df = df
if(!"rt_minutes" %in% colnames(FT$df)){
if(rtFormat == "sec"){
FT$df$rt_minutes <- df[,rtcol]/60
}else{
FT$df$rt_minutes <- df[,rtcol]
df[,rtcol] <- df[,rtcol]*60
}
}
FT$tablename = tablename
FT$editable = editable
FT$core = c(mzcol,rtcol)
FT$mz = mzcol
FT$rt = rtcol
FT$comments = commentcol
FT$fragments = fragmentcol
#default analysis result columns
FT$summaryStats = unname(unlist(sapply(c("topgroup","maxfold", "maxfoldover"),
grep,colnames(FT$df),
value = T)))
#make intensity value matrix
if(!is.null(anagrouptable)){
FT <- updateFTgrouping(FT,anagrouptable)
}
FT$others = colnames(FT$df)[which(!colnames(FT$df) %in% c(unlist(FT$gProps), unlist(FT$sProps),
FT$intensities, FT$intensities_norm,
FT$summaryStats, FT$commentcol,
FT$fragmentcol, FT$core))]
#selection and filtering for this object (these interact with selectize UI objects)
FT$sele = nrow(df)
FT$sortBy = mzcol
FT$sortByDecreasing = F
FT$order = nrow(df)
FT$filters = list( page = 1,
inpage = 1,
filters = list(Filter1 = list(selected = 1:nrow(df),
column = colnames(df)[1],
minSel = if(is.numeric(df[,colnames(df)[1]])){min(df[,colnames(df)[1]])}else{NULL},
maxSel = if(is.numeric(df[,colnames(df)[1]])){max(df[,colnames(df)[1]])}else{NULL},
modeSel = NULL,
txtSel = "",
active = F
)),
sele= 1:nrow(df),
sortBy = "mz",
sortByDecreasing = F,
order = 1:nrow(df)
)
FT$selectedCols = list(core = FT$core,
comments = FT$comments,
gProps = FT$gProps[[FT$gNames[1]]],
sProps = FT$sProps[[FT$gNames[1]]],
intensities = FT$anagroupnames[[FT$gNames[1]]],
others = FT$summaryStats)
FT$selectedGroup = FT$gNames[1]
FT$ctrlGroups = NULL
FT$useNorm = F
FT$.processHistory <- c(processHistory)
class(FT) <- "MseekFT"
return(FT)
}
#' updateDF
#'
#' Update dataframe b with data from dataframe a; must have equal number of rows
#' Existing data in b will be overridden if columns by the same name exist in a,
#' and retained otherwise.
#' Additional columns in a will be transferred to b.
#'
#' @param a source data.frame
#' @param b target data.frame
#'
#' @return Returns the updated data.frame b
#'
#' @export
updateDF <- function(a, b){
for (i in colnames(a)){
b[,i] <- a[,i, drop = FALSE]} #drop = FALSE added to avoid issues with list columns
return(b)
}
#' .getFTFormat
#'
#' Tries to identify the format of a feature table input data frame
#'
#' @param df data.frame
#'
#' @return Returns the format of the df object. Can be either of these:
#' \itemize{
#' \item "xcms PeakTable with Metaboseek Intensities"
#' \item "xcms Peak Table"
#' \item "mzMINE Aligned Peak List"
#' \item "unknown"
#' }
#'
#'
.getFTFormat <- function(df){
try({
if(all(c("rt", "mz") %in% colnames(df))
& any(grepl("\\.[Cc][Dd][Ff]__XIC$|\\.[Nn][Cc]__XIC$|\\.([Mm][Zz])?[Xx][Mm][Ll]__XIC$|\\.[Mm][Zz][Dd][Aa][Tt][Aa]__XIC$|\\.[Mm][Zz][Mm][Ll]__XIC$",
colnames(df)))){
return("xcms PeakTable with Metaboseek Intensities")
}
}, silent = TRUE)
try({
if(all(c("rt", "mz") %in% colnames(df))
& any(grepl("\\.[Cc][Dd][Ff]$|\\.[Nn][Cc]$|\\.([Mm][Zz])?[Xx][Mm][Ll]$|\\.[Mm][Zz][Dd][Aa][Tt][Aa]$|\\.[Mm][Zz][Mm][Ll]$",
colnames(df)))){
return("xcms Peak Table")
}
}, silent = TRUE)
try({
if(all(c("row m/z", "row retention time") %in% colnames(df))
& any(grepl("Peak area$",
colnames(df)))){
return("mzMINE Aligned Peak List")
}
}, silent = TRUE)
try({
if(all(c("Sample") %in% colnames(df))
&& all(is.character(unlist(df[1,])) | is.factor(unlist(df[1,])))
&& all(grepl("\\/",
df[-1,1]))
&& all(sapply(df[-1,which(!is.na(unlist(df[1,])))[-1]], function(x){is.numeric(type.convert(x))}))
){
return("MetaboAnalyst Peak Intensity Table")
}
}, silent = TRUE)
try({
if(c("Alignment ID") %in% df[,1]
&& c("Average Rt(min)") %in% df[,2]
&& c("Average Mz") %in% df[,3]
&& ("Class" %in% unlist(df[1,]) || "Class" %in% colnames(df))
&& all(is.character(unlist(df[1,])) | is.factor(unlist(df[1,])))
){
return("MS-DIAL Alignment Result")
}
}, silent = TRUE)
return("unknown")
}
#' .reformatFeatureTable
#'
#' Tries to identify the format of a feature table input data frame and
#' reformats it for compatibility with \code{constructFeatureTable}
#'
#' @param df data.frame
#' @param from input format. If auto, will run \code{.getFTFormat} to determine
#' the format.
#'
#' @return a list
#' \itemize{
#' \item df: reformatted input data.frame
#' \item grouping: anagrouptable format (if grouping information can be
#' extracted from input) or NULL
#' \item from: which input format was used
#' }
#'
#'
.reformatFeatureTable <- function(df,
from = c("auto",
"unknown",
"mzMINE Aligned Peak List",
"MetaboAnalyst Peak Intensity Table",
"MS-DIAL Alignment Result")){
from <- from[1]
if(from == "auto"){from <- .getFTFormat(df)}
res <- list(df = df,
grouping = NULL,
from = from)
rm(df) # to save memory
switch(from,
"unknown" = {},
"mzMINE Aligned Peak List" = {
res$grouping = data.frame(Column = grep("Peak area$",
colnames(res$df),
value = TRUE),
Group = "G1",
stringsAsFactors = FALSE)
res$df$mz <- res$df[["row m/z"]]
res$df$rt <- res$df[["row retention time"]]*60
#moving all intensity columns to the end
res$df <- res$df[,c(colnames(res$df)[!grepl("Peak area$",
colnames(res$df))],
colnames(res$df)[grepl("Peak area$",
colnames(res$df))])]
},
"MetaboAnalyst Peak Intensity Table" = {
res$grouping = data.frame(Column = colnames(res$df)[-1],
Group = unlist(res$df[1,])[-1],
stringsAsFactors = FALSE)
res$df <- data.frame(lapply(res$df[-1,], type.convert, as.is = TRUE),
stringsAsFactors = FALSE,
check.names = FALSE)
sp <- strsplit(as.character(res$df[,1]), split = '\\/')
res$df$mz <- as.numeric(sapply(sp,'[',1))
res$df$rt <- as.numeric(sapply(sp,'[',2))
},
"MS-DIAL Alignment Result" = {
if("Class" %in% unlist(res$df[1,])){
groups <- unlist(res$df[1, unlist(res$df[1,]) != ""])[-1]
}else{
groups <- colnames(res$df)[seq(which(colnames(res$df) == 'Class') + 1, ncol(res$df))]
}
colnames(res$df) <- unlist(res$df[which(res$df[[1]] == "Alignment ID"),])
res$df <- res$df[-seq_len(which(res$df[[1]] == "Alignment ID")+1),]
res$df <- data.frame(lapply(res$df, type.convert, as.is = TRUE),
stringsAsFactors = FALSE,
check.names = FALSE)
res$grouping = data.frame(Column = colnames(res$df)[seq(ncol(res$df) - length(groups) + 1, ncol(res$df))],
Group = groups,
stringsAsFactors = FALSE)
res$df[['mz']] <- res$df$`Average Mz`
res$df[['rt']] <- res$df$`Average Rt(min)`*60
}
)
return(res)
}
#' updateFeatureTable
#'
#' update an existing MseekFT with new data.frame columns (leaving intensitiy
#' column names and constants in effect, but updating information on which
#' "other" columns are present)
#'
#' @param FT MseekFT object
#' @param df source data.frame
#'
#' @return An updated \code{MseekFT} object
#'
#' @export
updateFeatureTable <- function(FT, df){
FT$df = updateDF(df,FT$df)
#if(!is.null(FT$anagrouptable)){
FT <- updateFTgrouping(FT,FT$anagrouptable)
# }
FT$summaryStats = unname(unlist(sapply(c("topgroup","maxfold", "maxfoldover"),
grep,colnames(FT$df),
value = T)))
FT$others = colnames(FT$df)[which(!colnames(FT$df) %in% c(unlist(FT$gProps), unlist(FT$sProps),
FT$intensities, FT$intensities_norm,
FT$summaryStats, FT$commentcol,
FT$fragmentcol, FT$core))]
FT$selectedCols = list(core = FT$core,
comments = FT$comments,
gProps = FT$gProps[[FT$gNames[1]]],
sProps = FT$sProps[[FT$gNames[1]]],
intensities = FT$anagroupnames[[FT$gNames[1]]],
others = FT$summaryStats)
FT$selectedGroup = FT$gNames[1]
return(FT)
}
#' updateFTgrouping
#'
#' update or construct grouping information of a MseekFT object
#'
#' @param FT MseekFT object
#' @param anagrouptable Analysis grouping table: a data.frame with columns
#' "Column" (containing column names from df with intensity values) and
#' "Group" (defining a group for each entry in "Column")
#'
#' @return An updated \code{MseekFT} object
#'
#' @export
updateFTgrouping <- function(FT,anagrouptable){
FT$anagrouptable = anagrouptable
if(!is.null(FT$anagrouptable) && nrow(FT$anagrouptable) >0 ){
newgrouping <- tableGrouping(FT$df,FT$anagrouptable)
FT$anagroupnames = newgrouping$anagroupnames
FT$intensities = unname(unlist(FT$anagroupnames))
FT$anagroupnums = newgrouping$anagroupnums
if(length(grep("__norm",colnames(FT$df)))>0){
FT$anagroupnames_norm = relist(paste0(unlist(FT$anagroupnames),
"__norm"),
FT$anagroupnames)
FT$intensities_norm = unname(unlist(FT$anagroupnames_norm))
}
#Group names
FT$gNames <- unique(anagrouptable$Group)
#Sample names, grouped
FT$sNames <- list()
for(i in names(FT$anagroupnames)){
#remove everything after double underscore (new default notation)
FT$sNames[[i]] <- gsub("__(.*)","",FT$anagroupnames[[i]])
#remove the XIC tag if it has only one underscore (old notation)
FT$sNames[[i]] <- gsub("_XIC","",FT$sNames[[i]])
#remove trailing underscores if any
FT$sNames[[i]] <- gsub("_$","",FT$sNames[[i]])
}
FT$gProps = list()
#detect and group property columns (group name + doubleunderscore)
for(i in FT$gNames){
#list all columns containing group names followed by double underscore
selCols <- grep(paste0(i,"__"),colnames(FT$df), value =T)
if(length(selCols)>0){
FT$gProps[[i]] <- selCols}
}
#detect and group sample property columns (sample name + doubleunderscore)
FT$sProps = list()
for(i in names(FT$gNames)){
#list all columns containing sample names from group i followed by double underscore which are not in the intensities table
selCols <- unlist(sapply(paste0(FT$sNames[[i]],"__"),
grep,
colnames(FT$df)[which(!colnames(FT$df) %in% c(FT$intensities,FT$intensities_norm))], value = T))
if(length(selCols)>0){
FT$sProps[[i]] <- sort(selCols)}
}
}
else{
FT$anagroupnames = NULL
FT$intensities = NULL
FT$anagroupnums = NULL
FT$anagroupnames_norm = NULL
FT$intensities_norm = NULL
FT$gNames = NULL
FT$sNames = NULL
FT$gProps = NULL
FT$sProps = NULL
}
return(FT)
}
#' updateFTgrouping
#'
#' update the FT index of the Mseek App (needed because this index is a named
#' list - extracts tablename entry from each MseekFT object in list)
#'
#' @param tables named list of MseekFT objects
#'
#' @return named list of MseekFT tablenames
#'
updateFTIndex <- function(tables){
out <- names(tables)
names(out) <- unname(sapply(tables, "[[", "tablename"))
return(out)
}
#' tableGrouping
#'
#' Groups a anagrouptable into a named list. Helper function to
#' reformat Table analysis grouping tables
#'
#' @param df feature table as data.frame, with intensity values.
#' @param anagrouptable Analysis grouping table: a data.frame with columns
#' \code{Column} (containing column names from df with intensity values) and
#' \code{Group} (defining a group for each entry in "Column")
#'
#' @return A named list, see \code{details}
#'
#' @details
#' Will return a named list with two list items
#' \itemize{
#' \item\code{anagroupnames} named list of intensity column names
#' \item\code{anagroupnums} named list of intensity column numbers
#' }
#'
tableGrouping <- function(df=NULL, anagrouptable){
## Make list object of grouped column names
if(is.null(anagrouptable$Column)){
ColumnNames <- gsub("-",".",paste0(basename(anagrouptable$File),"__XIC"))
ColumnNames[which(substring(ColumnNames,1,1) %in% as.character(0:9))] <- paste0("X",ColumnNames[which(substring(ColumnNames,1,1) %in% as.character(0:9))])
anagrouptable$Column <- ColumnNames
}
colme <- list()
for (l in unique(anagrouptable$Group)){
colme[[l]] <- as.character(anagrouptable$Column[which(anagrouptable$Group==l)])
}
anagroupnames <- colme
### Get column numbers from column names
if(!is.null(df)){
colnu <- integer(0)
for (i in anagrouptable$Column){
colnu<- c(colnu,which(colnames(df) == i))
}
## Make list object of grouped column numbers
colme <- list()
for (l in unique(anagrouptable$Group)){
colme[[l]] <- as.integer(colnu[which(anagrouptable$Group==l)])
}
anagroupnums <- colme
return(list(anagroupnames = anagroupnames,anagroupnums = anagroupnums))}
else{
return(list(anagroupnames = anagroupnames,anagroupnums = NULL))
}
}
#' tableWriter
#'
#' Writes a feature table to a file in a given format.
#'
#' @param df feature table as data.frame, with mz and rt columns.
#' @param fname file name
#' @param format output format
#' @param moreArgs named list of additional arguments to be
#' passed to \code{\link{asInclusionlist}} or \code{\link{asMetaboAnalyst}}
#'
#' @importFrom data.table fwrite
#'
#' @return Returns nothing, but writes a file to disk
#'
#' @examples
#' \dontrun{
#' MseekExamplePreload(data = FALSE, tables = TRUE)
#' tableWriter(tab1$df, "test_MA.csv", format = "MetaboAnalyst", moreArgs = list(groups = tab1$anagrouptable))
#' }
#'
#' @export
tableWriter <-function(df, fname, format = c("csv", "tsv",
"instrumentList",
"MetaboAnalyst"),
moreArgs = list()){
if(!length(format) || is.na(format[1])){
warning("File format selected was empty. Could not export data to file")
return(invisible(NULL))
}
tryCatch({
if(format[1] %in% c("csv","tsv")){
switch(format[1],
csv = {sep <- ","},
tsv = {sep <- "\t"})
#don't save complex columns
fwrite(df[,sapply(df,is.atomic)],
fname,
sep = sep,
quote = T,
row.names = F
)
return(invisible(NULL))
}
if(format[1] %in% c("instrumentList")){
moreArgs <- moreArgs[names(moreArgs) %in% c("rtwin", "polarity","instrument", "listType", "restrictRT")]
df <- do.call(asInclusionlist, c(list(parenttable = df[,sapply(df,is.atomic)]),
moreArgs))
fwrite(df,
fname,
sep = "\t",
quote = F,
row.names = F
)
return(fname)
}
if(format[1] %in% c("MetaboAnalyst")){
ma <- do.call(asMetaboAnalyst, c(list(df = df),
moreArgs))
fwrite(ma$header,
fname,
sep = ",",
quote = TRUE,
row.names = FALSE
)
fwrite(ma$body,
fname,
append = TRUE, #this also means colnames/header will not be written again
sep = ",",
quote = TRUE,
row.names = FALSE
)
return(invisible(NULL))
}
warning("Invalid file format selected. Could not export data to file")
},
error = function(e){
warning(paste0("An error occured while exporting a table: ", e))
})
}
#' asInclusionlist
#'
#' Reformats a feature table into a format that can be imported
#' into MS instruments as inclusion or exclusion lists.
#'
#' @param parenttable feature table as data.frame, with mz and rt columns.
#' @param rtwin retention time window (+/- df$rt) in seconds
#' @param polarity "Positive" or "Negative"
#' @param instrument which instrument is this table for?
#' @param listType "inclusion" or "exclusion"
#' @param restrictRT use retention time column from parenttable and rtwin to
#' restrict retention time in inclusion/exclusion list
#'
#' @return A data.frame that, when written to a file, has the expected format for a given instrument
#'
#'
#'
#' @export
asInclusionlist <- function(parenttable,
rtwin=2.5,
polarity = "Positive",
instrument = c("QExactive", "QExactive-HF"),
listType = "inclusion",
restrictRT = F){
if(listType == "exclusion"){
out <- data.frame("Mass [m/z]" = parenttable$mz,
"Formula [M]" = character(length(parenttable$mz)),
"Formula type" = character(length(parenttable$mz)),
Species = character(length(parenttable$mz)),
"CS [z]" = character(length(parenttable$mz)),
Polarity = as.character(rep(polarity,
length(parenttable$mz))),
"Start [min]" = if(!restrictRT){
character(length(parenttable$mz))
}else{as.numeric(round((parenttable$rt-rtwin)/60,3))},
"End [min]" = if(!restrictRT){
character(length(parenttable$mz))
}else{as.numeric(round((parenttable$rt+rtwin)/60,3))},
Comment = as.character(parenttable$comments),
stringsAsFactors = F)
colnames(out) <- c("Mass [m/z]","Formula [M]", "Formula type", "Species",
"CS [z]", "Polarity", "Start [min]",
"End [min]", "Comment")
}else{
out <- data.frame("Mass [m/z]" = parenttable$mz,
"Formula [M]" = character(length(parenttable$mz)),
"Formula type" = character(length(parenttable$mz)),
Species = character(length(parenttable$mz)),
"CS [z]" = character(length(parenttable$mz)),
Polarity = as.character(rep(polarity,
length(parenttable$mz))),
"Start [min]" = if(!restrictRT){
character(length(parenttable$mz))
}else{as.numeric(round((parenttable$rt-rtwin)/60,3))},
"End [min]" = if(!restrictRT){
character(length(parenttable$mz))
}else{as.numeric(round((parenttable$rt+rtwin)/60,3))},
"(N)CE" = character(length(parenttable$mz)),
"(N)CE type" = character(length(parenttable$mz)),
"MSX ID" = character(length(parenttable$mz)),
Comment = as.character(parenttable$comments),
stringsAsFactors = F)
colnames(out) <- c("Mass [m/z]","Formula [M]", "Formula type",
"Species", "CS [z]", "Polarity",
"Start [min]", "End [min]", "(N)CE", "(N)CE type",
"MSX ID", "Comment")
}
return(out)
}
#' asMetaboAnalyst
#'
#' Reformat a data.frame to MetaboAnalyst csv format
#'
#' @examples
#' \dontrun{
#' MseekExamplePreload(data = FALSE, tables = TRUE)
#' asMetaboAnalyst(tab1$df, groups = tab1$anagrouptable)
#' }
#'
#' @param df data.frame with intensity values and rt and mz columns
#' @param groups a data.frame with grouping information (Sample names in first column, groups in additional columns)
#'
#' @return A list with header and body of the MetaboAnalyst formatted table
#'
asMetaboAnalyst <- function(df, groups){
body <- data.frame(Sample = paste0(df$mz,"/",df$rt),
df[,groups[,1]],
stringsAsFactors = FALSE)
header <- as.data.frame(t(groups), stringsAsFactors = FALSE)
colnames(header) <- groups[,1]
header <- cbind(data.frame(Sample = colnames(groups[-1])),
header[-1,])
list(header = header,
body = body)
}
mjhelf/Mosaic documentation built on April 28, 2022, 11:32 a.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 asMetaboAnalyst asInclusionlist tableWriter tableGrouping updateFTIndex updateFTgrouping updateFeatureTable .reformatFeatureTable .getFTFormat updateDF constructFeatureTable
Documented in asInclusionlist asMetaboAnalyst constructFeatureTable .getFTFormat .reformatFeatureTable tableGrouping tableWriter updateDF updateFeatureTable updateFTgrouping updateFTIndex
#' constructFeatureTable
#'
#'
#' Constructor function for "MseekFT" object from a feature table data.frame.
#' uses and retains the original data frame plus names of columns containing
#' relevant data.
#'
#' NOTE: currently, only the default values for column names are supported.
#'
#' @param df feature table as data.frame, with retention time (rt), mz and
#' (optionally) intensity values.
#' @param mzcol column in df with mz values (columnname), defaults to "mz"
#' @param rtcol column in df with rt values (columnname), defaults to "rt"
#' @param commentcol column in df with comments (columnname), defaults to "comments"
#' @param fragmentcol column in df with fragmentation information (columnname),
#' defaults to "fragments"
#' @param rtFormat Are retention times given in seconds ("sec") or minutes ("min")
#' @param anagrouptable Analysis grouping table: a data.frame with columns
#' "Column" (containing column names from df with intensity values) and "Group"
#' (defining a group for each entry in "Column")
#' @param tablename Name of the table as displayed by Mseek
#' @param editable allow editing of this table in the Mseek app? if FALSE, only
#' comments column can be edited. editable tables are also not paginated.
#' @param processHistory a list of \code{\link[xcms]{processHistory}} objects
#'
#' @return an \code{MseekFT} object containing a feature table and metadata
#'
#' @export
constructFeatureTable <- function(df= data.frame(mz=numeric(3), rt = numeric(3)),# data frame
mzcol= "mz", #
rtcol= "rt", #column in df with mz values (columnname)
commentcol = "comments",
fragmentcol = "fragments",
rtFormat = "sec", # "sec" or "min"
anagrouptable = NULL,
tablename = "Custom Table",
editable = T,
processHistory = list()){ #T: free editing (add rows), but always see all columns in viewer, F: only comments can be edited directly, no adding of columns
#make columns if they don't exist:
if (class(try(df[,mzcol], silent = T))=="try-error"){ df[,mzcol] <- numeric(nrow(df)) }
if (class(try(df[,rtcol], silent = T))=="try-error"){ df[,rtcol] <- numeric(nrow(df)) }
if (class(try(df[,commentcol], silent = T))=="try-error"){ df[,commentcol] <- character(nrow(df)) }
if (class(try(df[,fragmentcol], silent = T))=="try-error"){ df[,fragmentcol] <- character(nrow(df)) }
FT = list()
#save the entire dataframe
FT$df = df
if(!"rt_minutes" %in% colnames(FT$df)){
if(rtFormat == "sec"){
FT$df$rt_minutes <- df[,rtcol]/60
}else{
FT$df$rt_minutes <- df[,rtcol]
df[,rtcol] <- df[,rtcol]*60
}
}
FT$tablename = tablename
FT$editable = editable
FT$core = c(mzcol,rtcol)
FT$mz = mzcol
FT$rt = rtcol
FT$comments = commentcol
FT$fragments = fragmentcol
#default analysis result columns
FT$summaryStats = unname(unlist(sapply(c("topgroup","maxfold", "maxfoldover"),
grep,colnames(FT$df),
value = T)))
#make intensity value matrix
if(!is.null(anagrouptable)){
FT <- updateFTgrouping(FT,anagrouptable)
}
FT$others = colnames(FT$df)[which(!colnames(FT$df) %in% c(unlist(FT$gProps), unlist(FT$sProps),
FT$intensities, FT$intensities_norm,
FT$summaryStats, FT$commentcol,
FT$fragmentcol, FT$core))]
#selection and filtering for this object (these interact with selectize UI objects)
FT$sele = nrow(df)
FT$sortBy = mzcol
FT$sortByDecreasing = F
FT$order = nrow(df)
FT$filters = list( page = 1,
inpage = 1,
filters = list(Filter1 = list(selected = 1:nrow(df),
column = colnames(df)[1],
minSel = if(is.numeric(df[,colnames(df)[1]])){min(df[,colnames(df)[1]])}else{NULL},
maxSel = if(is.numeric(df[,colnames(df)[1]])){max(df[,colnames(df)[1]])}else{NULL},
modeSel = NULL,
txtSel = "",
active = F
)),
sele= 1:nrow(df),
sortBy = "mz",
sortByDecreasing = F,
order = 1:nrow(df)
)
FT$selectedCols = list(core = FT$core,
comments = FT$comments,
gProps = FT$gProps[[FT$gNames[1]]],
sProps = FT$sProps[[FT$gNames[1]]],
intensities = FT$anagroupnames[[FT$gNames[1]]],
others = FT$summaryStats)
FT$selectedGroup = FT$gNames[1]
FT$ctrlGroups = NULL
FT$useNorm = F
FT$.processHistory <- c(processHistory)
class(FT) <- "MseekFT"
return(FT)
}
#' updateDF
#'
#' Update dataframe b with data from dataframe a; must have equal number of rows
#' Existing data in b will be overridden if columns by the same name exist in a,
#' and retained otherwise.
#' Additional columns in a will be transferred to b.
#'
#' @param a source data.frame
#' @param b target data.frame
#'
#' @return Returns the updated data.frame b
#'
#' @export
updateDF <- function(a, b){
for (i in colnames(a)){
b[,i] <- a[,i, drop = FALSE]} #drop = FALSE added to avoid issues with list columns
return(b)
}
#' .getFTFormat
#'
#' Tries to identify the format of a feature table input data frame
#'
#' @param df data.frame
#'
#' @return Returns the format of the df object. Can be either of these:
#' \itemize{
#' \item "xcms PeakTable with Metaboseek Intensities"
#' \item "xcms Peak Table"
#' \item "mzMINE Aligned Peak List"
#' \item "unknown"
#' }
#'
#'
.getFTFormat <- function(df){
try({
if(all(c("rt", "mz") %in% colnames(df))
& any(grepl("\\.[Cc][Dd][Ff]__XIC$|\\.[Nn][Cc]__XIC$|\\.([Mm][Zz])?[Xx][Mm][Ll]__XIC$|\\.[Mm][Zz][Dd][Aa][Tt][Aa]__XIC$|\\.[Mm][Zz][Mm][Ll]__XIC$",
colnames(df)))){
return("xcms PeakTable with Metaboseek Intensities")
}
}, silent = TRUE)
try({
if(all(c("rt", "mz") %in% colnames(df))
& any(grepl("\\.[Cc][Dd][Ff]$|\\.[Nn][Cc]$|\\.([Mm][Zz])?[Xx][Mm][Ll]$|\\.[Mm][Zz][Dd][Aa][Tt][Aa]$|\\.[Mm][Zz][Mm][Ll]$",
colnames(df)))){
return("xcms Peak Table")
}
}, silent = TRUE)
try({
if(all(c("row m/z", "row retention time") %in% colnames(df))
& any(grepl("Peak area$",
colnames(df)))){
return("mzMINE Aligned Peak List")
}
}, silent = TRUE)
try({
if(all(c("Sample") %in% colnames(df))
&& all(is.character(unlist(df[1,])) | is.factor(unlist(df[1,])))
&& all(grepl("\\/",
df[-1,1]))
&& all(sapply(df[-1,which(!is.na(unlist(df[1,])))[-1]], function(x){is.numeric(type.convert(x))}))
){
return("MetaboAnalyst Peak Intensity Table")
}
}, silent = TRUE)
try({
if(c("Alignment ID") %in% df[,1]
&& c("Average Rt(min)") %in% df[,2]
&& c("Average Mz") %in% df[,3]
&& ("Class" %in% unlist(df[1,]) || "Class" %in% colnames(df))
&& all(is.character(unlist(df[1,])) | is.factor(unlist(df[1,])))
){
return("MS-DIAL Alignment Result")
}
}, silent = TRUE)
return("unknown")
}
#' .reformatFeatureTable
#'
#' Tries to identify the format of a feature table input data frame and
#' reformats it for compatibility with \code{constructFeatureTable}
#'
#' @param df data.frame
#' @param from input format. If auto, will run \code{.getFTFormat} to determine
#' the format.
#'
#' @return a list
#' \itemize{
#' \item df: reformatted input data.frame
#' \item grouping: anagrouptable format (if grouping information can be
#' extracted from input) or NULL
#' \item from: which input format was used
#' }
#'
#'
.reformatFeatureTable <- function(df,
from = c("auto",
"unknown",
"mzMINE Aligned Peak List",
"MetaboAnalyst Peak Intensity Table",
"MS-DIAL Alignment Result")){
from <- from[1]
if(from == "auto"){from <- .getFTFormat(df)}
res <- list(df = df,
grouping = NULL,
from = from)
rm(df) # to save memory
switch(from,
"unknown" = {},
"mzMINE Aligned Peak List" = {
res$grouping = data.frame(Column = grep("Peak area$",
colnames(res$df),
value = TRUE),
Group = "G1",
stringsAsFactors = FALSE)
res$df$mz <- res$df[["row m/z"]]
res$df$rt <- res$df[["row retention time"]]*60
#moving all intensity columns to the end
res$df <- res$df[,c(colnames(res$df)[!grepl("Peak area$",
colnames(res$df))],
colnames(res$df)[grepl("Peak area$",
colnames(res$df))])]
},
"MetaboAnalyst Peak Intensity Table" = {
res$grouping = data.frame(Column = colnames(res$df)[-1],
Group = unlist(res$df[1,])[-1],
stringsAsFactors = FALSE)
res$df <- data.frame(lapply(res$df[-1,], type.convert, as.is = TRUE),
stringsAsFactors = FALSE,
check.names = FALSE)
sp <- strsplit(as.character(res$df[,1]), split = '\\/')
res$df$mz <- as.numeric(sapply(sp,'[',1))
res$df$rt <- as.numeric(sapply(sp,'[',2))
},
"MS-DIAL Alignment Result" = {
if("Class" %in% unlist(res$df[1,])){
groups <- unlist(res$df[1, unlist(res$df[1,]) != ""])[-1]
}else{
groups <- colnames(res$df)[seq(which(colnames(res$df) == 'Class') + 1, ncol(res$df))]
}
colnames(res$df) <- unlist(res$df[which(res$df[[1]] == "Alignment ID"),])
res$df <- res$df[-seq_len(which(res$df[[1]] == "Alignment ID")+1),]
res$df <- data.frame(lapply(res$df, type.convert, as.is = TRUE),
stringsAsFactors = FALSE,
check.names = FALSE)
res$grouping = data.frame(Column = colnames(res$df)[seq(ncol(res$df) - length(groups) + 1, ncol(res$df))],
Group = groups,
stringsAsFactors = FALSE)
res$df[['mz']] <- res$df$`Average Mz`
res$df[['rt']] <- res$df$`Average Rt(min)`*60
}
)
return(res)
}
#' updateFeatureTable
#'
#' update an existing MseekFT with new data.frame columns (leaving intensitiy
#' column names and constants in effect, but updating information on which
#' "other" columns are present)
#'
#' @param FT MseekFT object
#' @param df source data.frame
#'
#' @return An updated \code{MseekFT} object
#'
#' @export
updateFeatureTable <- function(FT, df){
FT$df = updateDF(df,FT$df)
#if(!is.null(FT$anagrouptable)){
FT <- updateFTgrouping(FT,FT$anagrouptable)
# }
FT$summaryStats = unname(unlist(sapply(c("topgroup","maxfold", "maxfoldover"),
grep,colnames(FT$df),
value = T)))
FT$others = colnames(FT$df)[which(!colnames(FT$df) %in% c(unlist(FT$gProps), unlist(FT$sProps),
FT$intensities, FT$intensities_norm,
FT$summaryStats, FT$commentcol,
FT$fragmentcol, FT$core))]
FT$selectedCols = list(core = FT$core,
comments = FT$comments,
gProps = FT$gProps[[FT$gNames[1]]],
sProps = FT$sProps[[FT$gNames[1]]],
intensities = FT$anagroupnames[[FT$gNames[1]]],
others = FT$summaryStats)
FT$selectedGroup = FT$gNames[1]
return(FT)
}
#' updateFTgrouping
#'
#' update or construct grouping information of a MseekFT object
#'
#' @param FT MseekFT object
#' @param anagrouptable Analysis grouping table: a data.frame with columns
#' "Column" (containing column names from df with intensity values) and
#' "Group" (defining a group for each entry in "Column")
#'
#' @return An updated \code{MseekFT} object
#'
#' @export
updateFTgrouping <- function(FT,anagrouptable){
FT$anagrouptable = anagrouptable
if(!is.null(FT$anagrouptable) && nrow(FT$anagrouptable) >0 ){
newgrouping <- tableGrouping(FT$df,FT$anagrouptable)
FT$anagroupnames = newgrouping$anagroupnames
FT$intensities = unname(unlist(FT$anagroupnames))
FT$anagroupnums = newgrouping$anagroupnums
if(length(grep("__norm",colnames(FT$df)))>0){
FT$anagroupnames_norm = relist(paste0(unlist(FT$anagroupnames),
"__norm"),
FT$anagroupnames)
FT$intensities_norm = unname(unlist(FT$anagroupnames_norm))
}
#Group names
FT$gNames <- unique(anagrouptable$Group)
#Sample names, grouped
FT$sNames <- list()
for(i in names(FT$anagroupnames)){
#remove everything after double underscore (new default notation)
FT$sNames[[i]] <- gsub("__(.*)","",FT$anagroupnames[[i]])
#remove the XIC tag if it has only one underscore (old notation)
FT$sNames[[i]] <- gsub("_XIC","",FT$sNames[[i]])
#remove trailing underscores if any
FT$sNames[[i]] <- gsub("_$","",FT$sNames[[i]])
}
FT$gProps = list()
#detect and group property columns (group name + doubleunderscore)
for(i in FT$gNames){
#list all columns containing group names followed by double underscore
selCols <- grep(paste0(i,"__"),colnames(FT$df), value =T)
if(length(selCols)>0){
FT$gProps[[i]] <- selCols}
}
#detect and group sample property columns (sample name + doubleunderscore)
FT$sProps = list()
for(i in names(FT$gNames)){
#list all columns containing sample names from group i followed by double underscore which are not in the intensities table
selCols <- unlist(sapply(paste0(FT$sNames[[i]],"__"),
grep,
colnames(FT$df)[which(!colnames(FT$df) %in% c(FT$intensities,FT$intensities_norm))], value = T))
if(length(selCols)>0){
FT$sProps[[i]] <- sort(selCols)}
}
}
else{
FT$anagroupnames = NULL
FT$intensities = NULL
FT$anagroupnums = NULL
FT$anagroupnames_norm = NULL
FT$intensities_norm = NULL
FT$gNames = NULL
FT$sNames = NULL
FT$gProps = NULL
FT$sProps = NULL
}
return(FT)
}
#' updateFTgrouping
#'
#' update the FT index of the Mseek App (needed because this index is a named
#' list - extracts tablename entry from each MseekFT object in list)
#'
#' @param tables named list of MseekFT objects
#'
#' @return named list of MseekFT tablenames
#'
updateFTIndex <- function(tables){
out <- names(tables)
names(out) <- unname(sapply(tables, "[[", "tablename"))
return(out)
}
#' tableGrouping
#'
#' Groups a anagrouptable into a named list. Helper function to
#' reformat Table analysis grouping tables
#'
#' @param df feature table as data.frame, with intensity values.
#' @param anagrouptable Analysis grouping table: a data.frame with columns
#' \code{Column} (containing column names from df with intensity values) and
#' \code{Group} (defining a group for each entry in "Column")
#'
#' @return A named list, see \code{details}
#'
#' @details
#' Will return a named list with two list items
#' \itemize{
#' \item\code{anagroupnames} named list of intensity column names
#' \item\code{anagroupnums} named list of intensity column numbers
#' }
#'
tableGrouping <- function(df=NULL, anagrouptable){
## Make list object of grouped column names
if(is.null(anagrouptable$Column)){
ColumnNames <- gsub("-",".",paste0(basename(anagrouptable$File),"__XIC"))
ColumnNames[which(substring(ColumnNames,1,1) %in% as.character(0:9))] <- paste0("X",ColumnNames[which(substring(ColumnNames,1,1) %in% as.character(0:9))])
anagrouptable$Column <- ColumnNames
}
colme <- list()
for (l in unique(anagrouptable$Group)){
colme[[l]] <- as.character(anagrouptable$Column[which(anagrouptable$Group==l)])
}
anagroupnames <- colme
### Get column numbers from column names
if(!is.null(df)){
colnu <- integer(0)
for (i in anagrouptable$Column){
colnu<- c(colnu,which(colnames(df) == i))
}
## Make list object of grouped column numbers
colme <- list()
for (l in unique(anagrouptable$Group)){
colme[[l]] <- as.integer(colnu[which(anagrouptable$Group==l)])
}
anagroupnums <- colme
return(list(anagroupnames = anagroupnames,anagroupnums = anagroupnums))}
else{
return(list(anagroupnames = anagroupnames,anagroupnums = NULL))
}
}
#' tableWriter
#'
#' Writes a feature table to a file in a given format.
#'
#' @param df feature table as data.frame, with mz and rt columns.
#' @param fname file name
#' @param format output format
#' @param moreArgs named list of additional arguments to be
#' passed to \code{\link{asInclusionlist}} or \code{\link{asMetaboAnalyst}}
#'
#' @importFrom data.table fwrite
#'
#' @return Returns nothing, but writes a file to disk
#'
#' @examples
#' \dontrun{
#' MseekExamplePreload(data = FALSE, tables = TRUE)
#' tableWriter(tab1$df, "test_MA.csv", format = "MetaboAnalyst", moreArgs = list(groups = tab1$anagrouptable))
#' }
#'
#' @export
tableWriter <-function(df, fname, format = c("csv", "tsv",
"instrumentList",
"MetaboAnalyst"),
moreArgs = list()){
if(!length(format) || is.na(format[1])){
warning("File format selected was empty. Could not export data to file")
return(invisible(NULL))
}
tryCatch({
if(format[1] %in% c("csv","tsv")){
switch(format[1],
csv = {sep <- ","},
tsv = {sep <- "\t"})
#don't save complex columns
fwrite(df[,sapply(df,is.atomic)],
fname,
sep = sep,
quote = T,
row.names = F
)
return(invisible(NULL))
}
if(format[1] %in% c("instrumentList")){
moreArgs <- moreArgs[names(moreArgs) %in% c("rtwin", "polarity","instrument", "listType", "restrictRT")]
df <- do.call(asInclusionlist, c(list(parenttable = df[,sapply(df,is.atomic)]),
moreArgs))
fwrite(df,
fname,
sep = "\t",
quote = F,
row.names = F
)
return(fname)
}
if(format[1] %in% c("MetaboAnalyst")){
ma <- do.call(asMetaboAnalyst, c(list(df = df),
moreArgs))
fwrite(ma$header,
fname,
sep = ",",
quote = TRUE,
row.names = FALSE
)
fwrite(ma$body,
fname,
append = TRUE, #this also means colnames/header will not be written again
sep = ",",
quote = TRUE,
row.names = FALSE
)
return(invisible(NULL))
}
warning("Invalid file format selected. Could not export data to file")
},
error = function(e){
warning(paste0("An error occured while exporting a table: ", e))
})
}
#' asInclusionlist
#'
#' Reformats a feature table into a format that can be imported
#' into MS instruments as inclusion or exclusion lists.
#'
#' @param parenttable feature table as data.frame, with mz and rt columns.
#' @param rtwin retention time window (+/- df$rt) in seconds
#' @param polarity "Positive" or "Negative"
#' @param instrument which instrument is this table for?
#' @param listType "inclusion" or "exclusion"
#' @param restrictRT use retention time column from parenttable and rtwin to
#' restrict retention time in inclusion/exclusion list
#'
#' @return A data.frame that, when written to a file, has the expected format for a given instrument
#'
#'
#'
#' @export
asInclusionlist <- function(parenttable,
rtwin=2.5,
polarity = "Positive",
instrument = c("QExactive", "QExactive-HF"),
listType = "inclusion",
restrictRT = F){
if(listType == "exclusion"){
out <- data.frame("Mass [m/z]" = parenttable$mz,
"Formula [M]" = character(length(parenttable$mz)),
"Formula type" = character(length(parenttable$mz)),
Species = character(length(parenttable$mz)),
"CS [z]" = character(length(parenttable$mz)),
Polarity = as.character(rep(polarity,
length(parenttable$mz))),
"Start [min]" = if(!restrictRT){
character(length(parenttable$mz))
}else{as.numeric(round((parenttable$rt-rtwin)/60,3))},
"End [min]" = if(!restrictRT){
character(length(parenttable$mz))
}else{as.numeric(round((parenttable$rt+rtwin)/60,3))},
Comment = as.character(parenttable$comments),
stringsAsFactors = F)
colnames(out) <- c("Mass [m/z]","Formula [M]", "Formula type", "Species",
"CS [z]", "Polarity", "Start [min]",
"End [min]", "Comment")
}else{
out <- data.frame("Mass [m/z]" = parenttable$mz,
"Formula [M]" = character(length(parenttable$mz)),
"Formula type" = character(length(parenttable$mz)),
Species = character(length(parenttable$mz)),
"CS [z]" = character(length(parenttable$mz)),
Polarity = as.character(rep(polarity,
length(parenttable$mz))),
"Start [min]" = if(!restrictRT){
character(length(parenttable$mz))
}else{as.numeric(round((parenttable$rt-rtwin)/60,3))},
"End [min]" = if(!restrictRT){
character(length(parenttable$mz))
}else{as.numeric(round((parenttable$rt+rtwin)/60,3))},
"(N)CE" = character(length(parenttable$mz)),
"(N)CE type" = character(length(parenttable$mz)),
"MSX ID" = character(length(parenttable$mz)),
Comment = as.character(parenttable$comments),
stringsAsFactors = F)
colnames(out) <- c("Mass [m/z]","Formula [M]", "Formula type",
"Species", "CS [z]", "Polarity",
"Start [min]", "End [min]", "(N)CE", "(N)CE type",
"MSX ID", "Comment")
}
return(out)
}
#' asMetaboAnalyst
#'
#' Reformat a data.frame to MetaboAnalyst csv format
#'
#' @examples
#' \dontrun{
#' MseekExamplePreload(data = FALSE, tables = TRUE)
#' asMetaboAnalyst(tab1$df, groups = tab1$anagrouptable)
#' }
#'
#' @param df data.frame with intensity values and rt and mz columns
#' @param groups a data.frame with grouping information (Sample names in first column, groups in additional columns)
#'
#' @return A list with header and body of the MetaboAnalyst formatted table
#'
asMetaboAnalyst <- function(df, groups){
body <- data.frame(Sample = paste0(df$mz,"/",df$rt),
df[,groups[,1]],
stringsAsFactors = FALSE)
header <- as.data.frame(t(groups), stringsAsFactors = FALSE)
colnames(header) <- groups[,1]
header <- cbind(data.frame(Sample = colnames(groups[-1])),
header[-1,])
list(header = header,
body = body)
}
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.