view BC/batch_correction_all_loess_wrapper.R @ 4:23314e1192d4 draft default tip

Uploaded
author melpetera
date Thu, 14 Jan 2021 09:56:58 +0000
parents
children
line wrap: on
line source

#!/usr/bin/env Rscript

library(batch) ## necessary for parseCommandArgs function

##------------------------------
## test help option
##------------------------------

# Prog. constants
argv.help <- commandArgs(trailingOnly = FALSE)
script.path <- sub("--file=", "", argv.help[grep("--file=", argv.help)])
prog.name <- basename(script.path)

# Test Help
if (length(grep('-h', argv.help)) > 0) {
  cat("Usage: Rscript ", 
    prog.name,
    "{args} \n",
    "parameters: \n",
    "\tdataMatrix {file}: set the input data matrix file (mandatory) \n",
    "\tsampleMetadata {file}: set the input sample metadata file (mandatory) \n",
    "\tvariableMetadata {file}: set the input variable metadata file (mandatory) \n",
    "\tmethod {opt}: set the method; can set to \"all_loess_pool\" or \"all_loess_sample\" (mandatory) \n",
    "\tspan {condition}: set the span condition; (mandatory) \n",
    "\tdataMatrix_out {file}: set the output data matrix file (mandatory) \n",
    "\tvariableMetadata_out {file}: set the output variable metadata file (mandatory) \n",
    "\tgraph_output {file}: set the output graph file (mandatory) \n",
    "\trdata_output {file}: set the output Rdata file (mandatory) \n",
    "\tbatch_col_name {val}: the column name for batch. Default value is \"batch\".\n",
    "\tinjection_order_col_name {val}: the column name for the injection order. Default value is \"injectionOrder\".\n",
    "\tsample_type_col_name {val}: the column name for the sample types. Default value is \"sampleType\".\n",
    "\tsample_type_tags {val}: the tags used inside the sample type column, defined as key/value pairs separated by commas (example: blank=blank,pool=pool,sample=sample).\n",
    "\n")
  quit(status = 0)
}

##------------------------------
## init. params
##------------------------------

args = parseCommandArgs(evaluate=FALSE) #interpretation of arguments given in command line as an R list of objects

# Set default col names
if ( ! 'batch_col_name' %in% names(args))
	args[['batch_col_name']] <- 'batch'
if ( ! 'injection_order_col_name' %in% names(args))
	args[['injection_order_col_name']] <- 'injectionOrder'
if ( ! 'sample_type_col_name' %in% names(args))
	args[['sample_type_col_name']] <- 'sampleType'
if ( ! 'sample_type_tags' %in% names(args))
	args[['sample_type_tags']] <- 'blank=blank,pool=pool,sample=sample'

# Parse sample type tags
sample.type.tags <- list()
for (kv in strsplit(strsplit(args$sample_type_tags, ',')[[1]], '='))
	sample.type.tags[[kv[[1]]]] <- kv[[2]]
if ( ! all(c('pool', 'blank', 'sample') %in% names(sample.type.tags)))
	stop("All tags pool, blank and sample must be defined in option sampleTypeTags.")
args$sample_type_tags <- sample.type.tags

##------------------------------
## init. functions
##------------------------------

source_local <- function(fname){
    argv <- commandArgs(trailingOnly = FALSE)
    base_dir <- dirname(substring(argv[grep("--file=", argv)], 8))
    source(paste(base_dir, fname, sep="/"))
}

## Import the different functions
source_local("batch_correction_all_loess_script.R")

argVc <- unlist(args)

##  argVc["method"] is either 'all_loess_pool' or 'all_loess_sample'
##  alternative version developped by CEA
##  all variables are treated with loess
##  the reference observations for loess are either 'pool'
## ('all_loess_pool') or 'sample' ('all_loess_sample')


##------------------------------
## Initializing
##------------------------------

## options
##--------

strAsFacL <- options()$stringsAsFactors
options(stringsAsFactors = FALSE)

## libraries
##----------

suppressMessages(library(ropls))

if(packageVersion("ropls") < "1.4.0")
    stop("Please use 'ropls' versions of 1.4.0 and above")

## constants
##----------

modNamC <- "Batch correction" ## module name

## log file
##---------

## sink(argVc["information"]) ## not implemented

cat("\nStart of the '", modNamC, "' Galaxy module call: ",
    format(Sys.time(), "%a %d %b %Y %X"), "\n", sep="")

## loading
##--------

rawMN <- t(as.matrix(read.table(argVc["dataMatrix"],
                                header = TRUE,
                                comment.char = '',
                                row.names = 1,
                                sep = "\t")))

samDF <- read.table(argVc["sampleMetadata"],
                    header = TRUE,
                    comment.char = '',
                    row.names = 1,
                    sep = "\t")

varDF <- read.table(argVc["variableMetadata"],
                    check.names = FALSE,
                    header = TRUE,
                    comment.char = '',
                    row.names = 1,
                    sep = "\t") ## not used; for compatibility only

refC <- tolower(gsub("all_loess_", "", argVc["method"]))

spnN <- as.numeric(argVc["span"])

## checking
##---------

stopifnot(refC %in% c('pool', 'sample'))
refC <- args$sample_type_tags[[refC]]

if(refC == args$sample_type_tags$pool &&
   !any(args$sample_type_tags$pool %in% samDF[, args$sample_type_col_name]))
    stop("No 'pool' found in the 'sampleType' column; use the samples as normalization reference instead")

refMN <- rawMN[samDF[, args$sample_type_col_name] == refC, ]
refNasZerVl <- apply(refMN, 2,
                     function(refVn)
                     all(sapply(refVn,
                                function(refN) {is.na(refN) || refN == 0})))

if(sum(refNasZerVl)) {

    refNasZerVi <- which(refNasZerVl)
    cat("The following variables have 'NA' or 0 values in all reference samples; they will be removed from the data:\n", sep = "")
    rawMN <- rawMN[, !refNasZerVl, drop = FALSE]
    varDF <- varDF[!refNasZerVl, , drop = FALSE]

}

##------------------------------
## Computation
##------------------------------


## ordering (batch and injection order)
##-------------------------------------

samDF[, "ordIniVi"] <- 1:nrow(rawMN)
ordBatInjVi <- order(samDF[, args$batch_col_name], samDF[, args$injection_order_col_name])
rawMN <- rawMN[ordBatInjVi, ]
samDF <- samDF[ordBatInjVi, ]

## signal drift and batch-effect correction
##-----------------------------------------

nrmMN <- shiftBatchCorrectF(rawMN,
                            samDF,
                            refC,
                            spnN)

## figure
##-------

cat("\nPlotting\n")

pdf(argVc["graph_output"], onefile = TRUE, width = 11, height = 7)
plotBatchF(rawMN, samDF, spnN)
plotBatchF(nrmMN, samDF, spnN)
dev.off()

## returning to initial order
##---------------------------

ordIniVi <- order(samDF[, "ordIniVi"])
nrmMN <- nrmMN[ordIniVi, ]
samDF <- samDF[ordIniVi, ]
samDF <- samDF[, colnames(samDF) != "ordIniVi", drop=FALSE]


##------------------------------
## Ending
##------------------------------


## saving
##-------

datMN <- nrmMN

datDF <- cbind.data.frame(dataMatrix = colnames(datMN),
                          as.data.frame(t(datMN)))
write.table(datDF,
            file = argVc["dataMatrix_out"],
            quote = FALSE,
            row.names = FALSE,
            sep = "\t")

varDF <- cbind.data.frame(variableMetadata = rownames(varDF),
                          varDF) ## not modified; for compatibility only
write.table(varDF,
            file = argVc["variableMetadata_out"],
            quote = FALSE,
            row.names = FALSE,
            sep = "\t")


res <- list(dataMatrix_raw = rawMN,
            dataMatrix_normalized = nrmMN,
            sampleMetadata = samDF)
save(res,
     file = argVc["rdata_output"]) ## for compatibility

## closing
##--------

cat("\nEnd of '", modNamC, "' Galaxy module call: ",
    as.character(Sys.time()), "\n", sep = "")

## sink()

options(stringsAsFactors = strAsFacL)


rm(argVc)