Merge pull request #40 from theislab/main

Merge from main

bin/bff.R

@@ -18,7 +18,7 @@ parser$add_argument("--methodsForConsensus", help='By default, a consensus call
 parser$add_argument("--cellbarcodeWhitelist", help='A vector of expected cell barcodes. This allows reporting on the total set of expected barcodes, not just those in the filtered count matrix',default=NULL)
 parser$add_argument("--metricsFile", help='If provided, summary metrics will be written to this file.', default="metrics_cell_hash_r.csv")
 parser$add_argument("--doTSNE", help='If true, tSNE will be run on the resulting hashing calls after each caller.', default=TRUE)
-parser$add_argument("--preprocess", help='If true, the PreProcess function by CellHashR is executed', default=FALSE)
+parser$add_argument("--preprocess_bff", help='If given, the PreProcess function by CellHashR is executed', action="store_true")
 parser$add_argument("--barcodeWhitelist", help='A vector of barcode names to retain, used for preprocessing step', default=NULL)
 parser$add_argument("--doHeatmap", help='f true, Seurat::HTOHeatmap will be run on the results of each caller.', default=TRUE)
 parser$add_argument("--perCellSaturation", help='An optional dataframe with the columns cellbarcode and saturation',default=NULL)
@@ -31,6 +31,7 @@ parser$add_argument("--assignmentOutBff", help="Prefix name for the file contain
 parser$add_argument("--outputdir", help='Output directory')
 args <- parser$parse_args()
 
+print(args$preprocess_bff)
 #Parameters originally Null
 methodsForConsensus <- args$methodsForConsensus
 if(is.null(methodsForConsensus)){
@@ -62,7 +63,7 @@ Argument <- c("HTO-File", "methods", "methodsForConsensus", "cellbarcodeWhitelis
 Value <- c(args$fileHto, args$methods, methodsForConsensus, cellbarcodeWhitelist, args$metricsFile, perCellSaturation, majorityConsensusThreshold, callerDisagreementThreshold, args$doTSNE, args$doHeatmap,args$chemistry)
 params <- data.frame(Argument, Value)
 
-if(as.logical(args$preprocess)){
+if(args$preprocess_bff == TRUE){
   print("Preprocessing activated")
   print(args$preprocess)
   #get barcodes
@@ -90,7 +91,7 @@ if (!is.null(args$methodsForConsensus)) {
 perCell_args <- args$perCellSaturation
 perCell <- ifelse(perCell_args == "null" || perCell_args == "Null", NULL, perCell_args)
 
-if(args$methodsForConsensus=="bff_raw" || args$methodsForConsensus=="bff_cluster" || args$methodsForConsensus=="bff_raw,bff_cluster" || is.null(args$methodsForConsensus)  )
+if(args$methodsForConsensus=="bff_raw" || args$methodsForConsensus=="bff_cluster" || args$methodsForConsensus=="bff_raw,bff_cluster" || args$methodsForConsensus=="bff_cluster,bff_raw"|| is.null(args$methodsForConsensus)  )
   #Only Bff in its different variations is available
   if (args$methods == "bff_raw") {
     print("Executing BFF raw")

modules/hash_demulti/bff.nf

@@ -29,13 +29,14 @@ process bff {
 
     script:
 
+        def run_preprocess = preprocess_bff != 'False' ? " --preprocess_bff" : ''
         """
         mkdir bff_${sampleId}
         bff.R --fileHto hto_data --methods $methods --methodsForConsensus $methodsForConsensus \
-                --cellbarcodeWhitelist $cellbarcodeWhitelist --metricsFile bff_${sampleId}_$metricsFile \
-                --doTSNE $doTSNE --doHeatmap $doHeatmap --perCellSaturation $perCellSaturation --majorityConsensusThreshold $majorityConsensusThreshold \
-                --chemistry $chemistry --callerDisagreementThreshold $callerDisagreementThreshold --outputdir bff_${sampleId} \
-                --assignmentOutBff $assignmentOutBff --preprocess $preprocess_bff --barcodeWhitelist $barcodeWhitelist
+        --cellbarcodeWhitelist $cellbarcodeWhitelist --metricsFile bff_${sampleId}_$metricsFile \
+        --doTSNE $doTSNE --doHeatmap $doHeatmap --perCellSaturation $perCellSaturation --majorityConsensusThreshold $majorityConsensusThreshold \
+        --chemistry $chemistry --callerDisagreementThreshold $callerDisagreementThreshold --outputdir bff_${sampleId} \
+        --assignmentOutBff $assignmentOutBff ${run_preprocess} --barcodeWhitelist $barcodeWhitelist
         """
 }
 

modules/hash_demulti/gmm_demux.nf

@@ -61,7 +61,7 @@ process gmm_demux{
 
 workflow gmm_demux_hashing{
 take: 
-        hto_matrix
+        input_list
   main:
         summary = params.summary
         report_gmm = params.report_gmm
@@ -70,7 +70,7 @@ take:
         threshold_gmm = params.threshold_gmm
         ambiguous = params.ambiguous
 
-        gmm_demux(hto_matrix,summary,report_gmm,mode,extract,threshold_gmm,ambiguous)
+        gmm_demux(input_list,summary,report_gmm,mode,extract,threshold_gmm,ambiguous)
 
   emit:
         gmm_demux.out.collect()

modules/hash_demultiplexing.nf

@@ -179,8 +179,12 @@ workflow hash_demultiplexing {
     if (params.gmmDemux == 'True') {
         Channel.fromPath(params.multi_input) \
                 | splitCsv(header:true) \
-                | map { row-> tuple(row.sampleId, params.hto_matrix_gmm_demux == 'raw' ? row.hto_matrix_raw : row.hto_matrix_filtered, row.hto_name_gmm) }
-                | gmm_demux_hashing
+
+                | map { row-> tuple(row.sampleId, 
+                        params.hto_matrix_gmm_demux == "raw" ? row.hto_matrix_raw : row.hto_matrix_filtered, 
+                        row.hto_name_gmm )}
+                | set {input_list_gmm_demux}
+                | gmm_demux_hashing(input_list_gmm_demux)
         gmmDemux_out = gmm_demux_hashing.out
     }
     else {

modules/single/hash_demulti/bff.nf

@@ -28,14 +28,15 @@ process bff {
         path "bff_${task.index}"
 
     script:
+        def run_preprocess = preprocess_bff != 'False' ? " --preprocess_bff" : ''
 
         """
         mkdir bff_${task.index}
         bff.R --fileHto hto_data  --methods $methods --methodsForConsensus $methodsForConsensus \
         --cellbarcodeWhitelist $cellbarcodeWhitelist --metricsFile bff_${task.index}_$metricsFile \
         --doTSNE $doTSNE --doHeatmap $doHeatmap --perCellSaturation $perCellSaturation --majorityConsensusThreshold $majorityConsensusThreshold \
         --chemistry $chemistry --callerDisagreementThreshold $callerDisagreementThreshold  --outputdir bff_${task.index} --assignmentOutBff $assignmentOutBff \
-        --preprocess $preprocess_bff --barcodeWhitelist $barcodeWhitelist
+        ${run_preprocess} --barcodeWhitelist $barcodeWhitelist
         """
 }
 

modules/single/hash_demulti/gmm_demux.nf

@@ -1,15 +1,17 @@
 #!/usr/bin/env nextflow
 nextflow.enable.dsl = 2
 
-process gmm_demux {
-    publishDir "$projectDir/$params.outdir/$sampleId/$params.mode/hash_demulti/gmm_demux", mode:'copy'
+
+process gmm_demux{
+    publishDir "$projectDir/$params.outdir/$params.mode/hash_demulti/gmm_demux", mode:'copy'
+
     label 'small_mem'
     conda "$projectDir/conda/gmm_demux.yml"
 
     input:
-        tuple val(sampleId), path(filtered_hto_matrix_dir), val(hto_name_gmm)
+        path filtered_hto_matrix_dir
         //HTO names as string separated by commas
-        //val hto_name_gmm
+        val hto_name_gmm
         //mode 2
         //need estimate number of cells in the single cell assay
         //obligatory
@@ -26,29 +28,31 @@ process gmm_demux {
         val ambiguous
 
     output:
-        path "gmm_demux_${sampleId}"
+
+        path "gmm_demux_${task.index}"
+
 
     script:
         def extract_droplets = extract != 'None' ? " -x ${extract}" : ''
         def ambiguous_droplets = extract != 'None' ? " --ambiguous ${ambiguous}" : ''
 
         if (mode_GMM == 'csv') {
             """
-            mkdir gmm_demux_${sampleId}
-            touch gmm_demux_${sampleId}_$report_gmm
-
-            GMM-demux -c $filtered_hto_matrix_dir $hto_name_gmm -u $summary --report gmm_demux_${sampleId}_$report_gmm --full gmm_demux_${sampleId} $extract_droplets -t $threshold_gmm
-            gmm_demux_params.py --path_hto $filtered_hto_matrix_dir --hto_name_gmm $hto_name_gmm --summary $summary --report gmm_demux_${sampleId}_$report_gmm   --mode $mode_GMM  $extract_droplets --threshold_gmm $threshold_gmm $ambiguous_droplets  --outputdir gmm_demux_${sampleId}
-
+            mkdir gmm_demux_${task.index}
+            touch gmm_demux_${task.index}_$report_gmm
+            
+            GMM-demux -c $filtered_hto_matrix_dir $hto_name_gmm -u $summary --report gmm_demux_${task.index}_$report_gmm --full gmm_demux_${task.index} $extract_droplets -t $threshold_gmm
+            gmm_demux_params.py --path_hto $filtered_hto_matrix_dir --hto_name_gmm $hto_name_gmm --summary $summary --report gmm_demux_${task.index}_$report_gmm   --mode $mode_GMM  $extract_droplets --threshold_gmm $threshold_gmm $ambiguous_droplets  --outputdir gmm_demux_${task.index}
+            
             """
         }else {
             """
-            mkdir gmm_demux_${sampleId}
-            touch gmm_demux_${sampleId}_$report_gmm
-
-            GMM-demux $filtered_hto_matrix_dir $hto_name_gmm -u $summary -r gmm_demux_${sampleId}_$report_gmm --full gmm_demux_${sampleId} -o gmm_demux_${sampleId} $extract_droplets -t $threshold_gmm
-            gmm_demux_params.py --path_hto $filtered_hto_matrix_dir --hto_name_gmm $hto_name_gmm --summary $summary --report gmm_demux_${sampleId}_$report_gmm --mode $mode_GMM $extract_droplets  --threshold_gmm $threshold_gmm $ambiguous_droplets --outputdir gmm_demux_${sampleId}
-
+            mkdir gmm_demux_${task.index}
+            touch gmm_demux_${task.index}_$report_gmm
+            
+            GMM-demux $filtered_hto_matrix_dir $hto_name_gmm -u $summary -r gmm_demux_${task.index}_$report_gmm --full gmm_demux_${task.index} -o gmm_demux_${task.index} $extract_droplets -t $threshold_gmm
+            gmm_demux_params.py --path_hto $filtered_hto_matrix_dir --hto_name_gmm $hto_name_gmm --summary $summary --report gmm_demux_${task.index}_$report_gmm --mode $mode_GMM $extract_droplets  --threshold_gmm $threshold_gmm $ambiguous_droplets --outputdir gmm_demux_${task.index}
+            
             """
         }
 }
@@ -57,15 +61,17 @@ workflow gmm_demux_hashing {
 take:
         hto_matrix
   main:
+        hto_name_gmm = params.hto_name_gmm
         summary = params.summary
         report_gmm = params.report_gmm
         mode = params.mode_GMM
         extract = params.extract
         threshold_gmm = params.threshold_gmm
         ambiguous = params.ambiguous
 
-        gmm_demux(hto_matrix, summary, report_gmm, mode, extract, threshold_gmm, ambiguous)
 
+        gmm_demux(hto_matrix,hto_name_gmm,summary,report_gmm,mode,extract,threshold_gmm,ambiguous)
+
   emit:
         gmm_demux.out.collect()
 }

test.config

@@ -1,8 +1,8 @@
 params {
   outdir = "result_test"
   // input for hashing-based deconvolution
-  hto_matrix_raw = "$projectDir/test_data/hto"
-  hto_matrix_filtered = "$projectDir/test_data/hto"
+  hto_matrix_raw = "$projectDir/test_data/hto_gz"
+  hto_matrix_filtered = "$projectDir/test_data/hto_gz"
   rna_matrix_raw = "$projectDir/test_data/rna"
   rna_matrix_filtered = "$projectDir/test_data/rna"
 
@@ -23,4 +23,7 @@ params {
   // donor genotype file provided by popscle doesnt work on souporcell
   use_known_genotype = "False"
   ignore = "True"
+  //gmm_demux input
+  hto_name_gmm = "hto_1,hto_2"
+
 }