refactor: use consistent config file within DNAm QC pipeline

array/DNAm/preprocessing/CETYGOdeconvolution.r

@@ -21,10 +21,10 @@
 #----------------------------------------------------------------------#
 
 args<-commandArgs(trailingOnly = TRUE)
-dataDir <- args[1]
+dataDir <- args[[1]]
+configFile <- args[[2]]
 
 gdsFile <-paste0(dataDir, "/2_gds/raw.gds")
-configFile <- paste0(dataDir, "/config.r")
 
 source(configFile)
 
@@ -50,17 +50,17 @@
 setwd(dataDir)
 
 # load sample sheet
 sampleSheet<-read.csv("0_metadata/sampleSheet.csv", na.strings = c("", "NA"), stringsAsFactors = FALSE)
 # if no column Basename, creates from columns Chip.ID and Chip.Location
 if(!"Basename" %in% colnames(sampleSheet)){
 	sampleSheet$Basename<-paste(sampleSheet$Chip.ID, sampleSheet$Chip.Location, sep = "_")
 }
 sampleSheet$Cell_Type <- as.factor(sampleSheet$Cell_Type)
 
 
 gfile<-openfn.gds(gdsFile, readonly = FALSE, allow.fork = TRUE)
 # ensure sample sheet is in same order as data
 sampleSheet<-sampleSheet[match(colnames(gfile), sampleSheet$Basename),]
 # extract a few useful matrices
 if(arrayType == "V2"){
 	 rawbetas<-epicv2clean(betas(gfile)[])

array/DNAm/preprocessing/QC.rmd

@@ -23,16 +23,15 @@ library(RColorBrewer, warn.conflicts = FALSE, quietly = TRUE)
 library(pheatmap, warn.conflicts = FALSE, quietly = TRUE)
 library(data.table, warn.conflicts = FALSE, quietly = TRUE)
 
-
-source(args[4]) ### change the content of this file to run QC with different thresholds
-### prior to running this Rmarkdown which summarises the QC output, QC metrics must have been generated
-
-dataDir <- args[2]
-refDir <- args[3]
-most_recent_git_tag <- args[6]
-current_commit_hash <- args[7]
+dataDir <- args[[2]]
+refDir <- args[[3]]
+configFile <- args[[4]]
+most_recent_git_tag <- args[[6]]
+current_commit_hash <- args[[7]]
 setwd(dataDir) 
 
+source(configFile)
+
 qcData <-paste0(dataDir, "/2_gds/QCmetrics/QCmetrics.rdata")
 
 load(qcData)

array/DNAm/preprocessing/calcQCMetrics.r

@@ -21,13 +21,13 @@
 # DEFINE PARAMETERS
 #----------------------------------------------------------------------#
 args<-commandArgs(trailingOnly = TRUE)
-dataDir <- args[1]
-refDir <- args[2]
+dataDir <- args[[1]]
+refDir <- args[[2]]
+configFile <- args[[3]]
 
 gdsFile <-paste0(dataDir, "/2_gds/raw.gds")
 qcData <-paste0(dataDir, "/2_gds/QCmetrics/QCmetrics.rdata")
 genoFile <- paste0(dataDir, "/0_metadata/epicSNPs.raw")
-configFile <- paste0(dataDir, "/config.r")
 
 source(configFile)
 

array/DNAm/preprocessing/checkColnamesSampleSheet.r

@@ -1,80 +1,78 @@
-
-
 ## This script checks that sample sheet columns are formatted correctly prior to DNAm QC ##
 
 
 args <- commandArgs(trailingOnly = TRUE)
-dataDir <- args[1]
-configFile <- paste0(dataDir, "/config.r")
+configFile <- args[[1]]
 
 # Load libraries
 library(stringdist, warn.conflicts = FALSE, quietly = TRUE) # for amatch()
-'%ni%' <- Negate('%in%') # define '%ni%' (not in)
+"%ni%" <- Negate("%in%") # define '%ni%' (not in)
 
 # Load sample sheet
 sampleSheet <- read.csv(paste0(dataDir, "/0_metadata/sampleSheet.csv"), na.strings = c("", "NA"), stringsAsFactors = FALSE)
 
 # Column names to test
-req_cols <- c('Sample_ID','Individual_ID') 												# required column names
-bsnm_cols <- c('Basename','Chip_ID','Chip_Location','Sentrix_ID','Sentrix_Position')	# required when Basename not present
-opt_cols <- c('Age')																	# optional column names
-cond_cols <- c('Sex','Genotype_IID','Cell_Type')										# conditionally column names
+req_cols <- c("Sample_ID", "Individual_ID") # required column names
+bsnm_cols <- c("Basename", "Chip_ID", "Chip_Location", "Sentrix_ID", "Sentrix_Position") # required when Basename not present
+opt_cols <- c("Age") # optional column names
+cond_cols <- c("Sex", "Genotype_IID", "Cell_Type") # conditionally column names
 
 # source checkColnames()
 source("checkColnamesFunction.r")
 
 
-#1. Check required column names ------------------------------------------------------------------------------
+# 1. Check required column names ------------------------------------------------------------------------------
 cat("1. Checking required column names: ")
-cat(c('Basename',req_cols,'\n'))
+cat(c("Basename", req_cols, "\n"))
 
 # check Basename first
-b <- checkColnames(sampleSheet, bsnm_cols[1], type='Required', verbose=F)
-
-if(b$allPresent){
-	# if Basename present, continue to check Basename alongside other required columns
-	checkColnames(sampleSheet, c('Basename',req_cols), type='Required')
-}else{
-	# if Basename not present, check Chip and Sentrix as alternatives
-	chip <- checkColnames(sampleSheet, bsnm_cols[2:3], type='Required', verbose=F)
-	sntrx <- checkColnames(sampleSheet, bsnm_cols[4:5], type='Required', verbose=F)
-	
-	# if either Chip or Sentrix present, continue to check other required columns
-	if(any(chip$allPresent | sntrx$allPresent)){
-		cat("Basename column not found, but at least 2 of the following alternative columns are present: ", bsnm_cols[2:5],'\n')
-		cat("Checking remaining required columns",'\n')
-		checkColnames(sampleSheet, req_cols, type='Required')
-	}else{
-		cat("Basename column not found, and neither set of alternative column names are present: ", bsnm_cols[2:5],'\n')
-		cat("Checking remaining required columns",'\n')
-		checkColnames(sampleSheet, req_cols, type='Required')
-	}
+b <- checkColnames(sampleSheet, bsnm_cols[1], type = "Required", verbose = F)
+
+if (b$allPresent) {
+    # if Basename present, continue to check Basename alongside other required columns
+    checkColnames(sampleSheet, c("Basename", req_cols), type = "Required")
+} else {
+    # if Basename not present, check Chip and Sentrix as alternatives
+    chip <- checkColnames(sampleSheet, bsnm_cols[2:3], type = "Required", verbose = F)
+    sntrx <- checkColnames(sampleSheet, bsnm_cols[4:5], type = "Required", verbose = F)
+
+    # if either Chip or Sentrix present, continue to check other required columns
+    if (any(chip$allPresent | sntrx$allPresent)) {
+        cat("Basename column not found, but at least 2 of the following alternative columns are present: ", bsnm_cols[2:5], "\n")
+        cat("Checking remaining required columns", "\n")
+        checkColnames(sampleSheet, req_cols, type = "Required")
+    } else {
+        cat("Basename column not found, and neither set of alternative column names are present: ", bsnm_cols[2:5], "\n")
+        cat("Checking remaining required columns", "\n")
+        checkColnames(sampleSheet, req_cols, type = "Required")
+    }
 }
 
 
 
 
-#2. Check optional column names ------------------------------------------------------------------------------
+# 2. Check optional column names ------------------------------------------------------------------------------
 cat("2. Checking optional column names: ")
-cat(opt_cols,'\n')
-checkColnames(sampleSheet, opt_cols, type='Optional')
+cat(opt_cols, "\n")
+checkColnames(sampleSheet, opt_cols, type = "Optional")
 
 
-#3. Check conditional column names ---------------------------------------------------------------------------
-cat("Sourcing conditional variables from config.r",'\n')
+# 3. Check conditional column names ---------------------------------------------------------------------------
+cat("Sourcing conditional variables from config.r", "\n")
 source(configFile)
-cond_status <- c(sexCheck,snpCheck,ctCheck) # T/Fs from config file
-cat(paste0(c("sexCheck","snpCheck","ctCheck"),"=", cond_status),'\n')
+cond_status <- c(sexCheck, snpCheck, ctCheck) # T/Fs from config file
+cat(paste0(c("sexCheck", "snpCheck", "ctCheck"), "=", cond_status), "\n")
 
 # subset conditional colnames to those TRUE in config
 cond_cols.filtered <- cond_cols[cond_status]
 
-if(all(cond_status==F)){
-	cat("No conditional variables to check")
-}else{
-	cat("3. Checking conditional column names: ",'\n')
-	cat(cond_cols.filtered,'\n')
-	checkColnames(sampleSheet, cond_cols.filtered, type='Conditional')
+if (all(cond_status == F)) {
+    cat("No conditional variables to check")
+} else {
+    cat("3. Checking conditional column names: ", "\n")
+    cat(cond_cols.filtered, "\n")
+    checkColnames(sampleSheet, cond_cols.filtered, type = "Conditional")
 }
 
-# ---------------------------------------------------------------------------------------------------------- #
+# ---------------------------------------------------------------------------------------------------------- #
+

array/DNAm/preprocessing/checkRconfigFile.r

@@ -22,8 +22,7 @@
 print("checking config.r file parameters are present and correctly formatted...")
 
 args <- commandArgs(trailingOnly = TRUE)
-dataDir <- args[1]
-configFile <- file.path(dataDir, "config.r")
+configFile <- args[[1]]
 
 source(configFile)
 

array/DNAm/preprocessing/clusterCellTypes.r

@@ -18,14 +18,14 @@
 #----------------------------------------------------------------------#
 
 args<-commandArgs(trailingOnly = TRUE)
-dataDir <- args[1]
-refDir <- args[2]
+dataDir <- args[[1]]
+refDir <- args[[2]]
+configFile <- args[[3]]
 
 gdsFile <-paste0(dataDir, "/2_gds/raw.gds")
 qcOutFolder<-paste0(dataDir, "/2_gds/QCmetrics")
 qcData <-paste0(dataDir, "/2_gds/QCmetrics/QCmetrics.rdata")
 genoFile <- paste0(dataDir, "/0_metadata/epicSNPs.raw")
-configFile <- paste0(dataDir, "/config.r")
 
 source(configFile)
 

array/DNAm/preprocessing/jobSubmission/1_runDNAmQC.sh

@@ -52,7 +52,7 @@ module load $RVERS   # load specified R version
 
 cd ${SCRIPTSDIR}/array/DNAm/preprocessing/
 
-Rscript checkRconfigFile.r ${DATADIR}
+Rscript checkRconfigFile.r ${RCONFIG}
 config_malformed=$?
 if [[ "${config_malformed}" -ne 0 ]]; then
     print_error_message \
@@ -71,7 +71,7 @@ if [[ "${library_did_not_install}" -ne 0 ]]; then
         "Exiting..."
 fi
 
-Rscript checkColnamesSampleSheet.r ${DATADIR}
+Rscript checkColnamesSampleSheet.r ${RCONFIG}
 sample_sheet_malformed=$?
 if [[ "${sample_sheet_malformed}" -ne 0 ]]; then
     print_error_message \
@@ -81,7 +81,7 @@ fi
 
 mkdir -p ${GDSDIR}/QCmetrics
 
-Rscript loadDataGDS.r ${DATADIR}
+Rscript loadDataGDS.r ${DATADIR} ${RCONFIG}
 gds_problem_identified=$?
 if [[ "${gds_problem_identified}" -ne 0 ]]; then
     print_error_message \
@@ -91,9 +91,9 @@ fi
 
 chmod 755 ${DATADIR}/2_gds/raw.gds
 
-Rscript calcQCMetrics.r ${DATADIR} ${REFDIR}
+Rscript calcQCMetrics.r ${DATADIR} ${REFDIR} ${RCONFIG}
 
-Rscript clusterCellTypes.r ${DATADIR} ${REFDIR}
+Rscript clusterCellTypes.r ${DATADIR} ${REFDIR} ${RCONFIG}
 
 most_recent_git_tag=$(git describe --tags --abbrev=0)
 current_commit_hash=$(git rev-parse HEAD)
@@ -106,12 +106,12 @@ mv QC.html ${GDSDIR}/QCmetrics/
 
 mkdir -p ${DATADIR}/3_normalised
 
-Rscript normalisation.r ${DATADIR} ${REFDIR}
+Rscript normalisation.r ${DATADIR} ${REFDIR} ${RCONFIG}
 chmod 755 ${DATADIR}/2_gds/rawNorm.gds
 
 mkdir -p ${GDSDIR}/QCmetrics/CETYGO
 
-Rscript CETYGOdeconvolution.r ${DATADIR}
+Rscript CETYGOdeconvolution.r ${DATADIR} ${RCONFIG}
 
 ## print finish date and time
 echo Job finished on:

array/DNAm/preprocessing/loadDataGDS.r

@@ -17,11 +17,11 @@
 # DEFINE PARAMETERS
 #----------------------------------------------------------------------#
 args <- commandArgs(trailingOnly = TRUE)
-dataDir <- args[1]
+dataDir <- args[[1]]
+configFile <- args[[2]]
 
 gdsFile <- file.path(dataDir, "2_gds/raw.gds")
 
-configFile <- paste0(dataDir, "/config.r")
 source(configFile)
 
 arrayType <- toupper(arrayType)

array/DNAm/preprocessing/normalisation.r

@@ -19,13 +19,14 @@
 # DEFINE PARAMETERS
 #----------------------------------------------------------------------#
 args<-commandArgs(trailingOnly = TRUE)
-dataDir <- args[1]
-refDir <- args[2]
+dataDir <- args[[1]]
+refDir <- args[[2]]
+configFile <- args[[3]]
+
 gdsFile <-file.path(dataDir, "/2_gds/raw.gds")
 normgdsFile<-sub("\\.gds", "Norm.gds", gdsFile)
 qcOutFolder<-file.path(dataDir, "/2_gds/QCmetrics")
 normData<-file.path(dataDir, "/3_normalised/normalised.rdata")
-configFile <- paste0(dataDir, "/config.r")
 
 source(configFile)