soupx.R

#!/usr/bin/Rscript

library(SoupX)

#run with one positional argument - the citeseq combined sample identifier
#(that cellranger uses to create its output folder)
args = commandArgs(TRUE)
dataDir = file.path(args[1],'outs')
#load just the CITE part
sc = load10X(dataDir, includeFeatures=c('Antibody Capture'))
#this might sometimes actually fail. so try() it just in case
soupQuantile = 0.25
tfidfMin = 0.2
message = try(sc <- autoEstCont(sc, soupQuantile=soupQuantile, tfidfMin=tfidfMin, forceAccept=TRUE))
#we really don't want this to fail though. so keep dropping the parameters and trying again
#(once tfidMin hits 0.05 the next iteration puts it at 0, so just give up)
while ((class(message) == 'try-error') && (tfidfMin > 0.05))
{
	soupQuantile = soupQuantile - 0.05
	tfidfMin = tfidfMin - 0.05
	message = try(sc <- autoEstCont(sc, soupQuantile=soupQuantile, tfidfMin=tfidfMin, forceAccept=TRUE))
}
if (class(message) != 'try-error')
{
	out = adjustCounts(sc)
	#create soupx-cite subfolder within filtered_feature_bc_matrix
	#and write out a cellranger3-formatted mtx export there
	featurePath = file.path(dataDir,'filtered_feature_bc_matrix')
	soupxPath = file.path(featurePath,'soupx-cite')
	dir.create(soupxPath)
	Matrix::writeMM(out, file.path(soupxPath,'matrix.mtx'))
	system(paste('gzip', file.path(soupxPath,'matrix.mtx')))
	system(paste('cp', file.path(featurePath,'barcodes.tsv.gz'), file.path(soupxPath,'barcodes.tsv.gz')))
	#the genes need a little postprocessing nudge - filter it to just the AB_ ones, then compress
	system(paste('zcat', file.path(featurePath,'features.tsv.gz'), '| grep "Antibody Capture" >', file.path(soupxPath,'features.tsv')))
	system(paste('gzip', file.path(soupxPath,'features.tsv')))
}