Add option to auto-detect strandedness

drpatelh · Mar 3, 2022 · 0d760d2 · 0d760d2
1 parent 9b1cc4b
commit 0d760d2
Show file tree

Hide file tree

Showing 5 changed files with 52 additions and 8 deletions.
diff --git a/lib/WorkflowRnaseq.groovy b/lib/WorkflowRnaseq.groovy
@@ -2,6 +2,8 @@
 // This file holds several functions specific to the workflow/rnaseq.nf in the nf-core/rnaseq pipeline
 //
 
+import groovy.json.JsonSlurper
+
 class WorkflowRnaseq {
 
     //
@@ -152,6 +154,24 @@ class WorkflowRnaseq {
         }
     }
 
+    //
+    // Function that parses Salmon quant 'meta_info.json' output file to get inferred strandedness
+    //
+    public static String getSalmonInferredStrandedness(json_file) {
+        def lib_type = new JsonSlurper().parseText(json_file.text).get('library_types')[0]
+        def strandedness = 'reverse'
+        if (lib_type) {
+            if (lib_type in ['U', 'IU']) {
+                strandedness = 'unstranded'
+            } else if (lib_type in ['SF', 'ISF']) {
+                strandedness = 'forward'
+            } else if (lib_type in ['SR', 'ISR']) {
+                strandedness = 'reverse'
+            }
+        }
+        return strandedness
+    }
+
     //
     // Function that parses and returns the alignment rate from the STAR log output
     //

diff --git a/modules/nf-core/modules/salmon/quant/main.nf b/modules/nf-core/modules/salmon/quant/main.nf
diff --git a/modules/nf-core/modules/salmon/quant/meta.yml b/modules/nf-core/modules/salmon/quant/meta.yml
diff --git a/subworkflows/local/quantify_salmon.nf b/subworkflows/local/quantify_salmon.nf
@@ -55,14 +55,15 @@ workflow QUANTIFY_SALMON {
         SALMON_TX2GENE.out.tsv.collect()
     )
 
-    SALMON_SE_TRANSCRIPT (
-        SALMON_TXIMPORT.out.counts_transcript,
-        SALMON_TXIMPORT.out.tpm_transcript,
-        SALMON_TX2GENE.out.tsv.collect()
-    )
+    // SALMON_SE_TRANSCRIPT (
+    //     SALMON_TXIMPORT.out.counts_transcript,
+    //     SALMON_TXIMPORT.out.tpm_transcript,
+    //     SALMON_TX2GENE.out.tsv.collect()
+    // )
 
     emit:
     results                       = SALMON_QUANT.out.results                      // channel: [ val(meta), results_dir ]
+    json_info                     = SALMON_QUANT.out.json_info                    // channel: [ val(meta), json_info ]
 
     tpm_gene                      = SALMON_TXIMPORT.out.tpm_gene                  // channel: [ val(meta), counts ]
     counts_gene                   = SALMON_TXIMPORT.out.counts_gene               // channel: [ val(meta), counts ]
@@ -77,7 +78,7 @@ workflow QUANTIFY_SALMON {
 
     merged_counts_transcript      = SALMON_TXIMPORT.out.counts_transcript         //    path: *.transcript_counts.tsv
     merged_tpm_transcript         = SALMON_TXIMPORT.out.tpm_transcript            //    path: *.transcript_tpm.tsv
-    merged_transcript_rds         = SALMON_SE_TRANSCRIPT.out.rds                  //    path: *.rds
+    // merged_transcript_rds         = SALMON_SE_TRANSCRIPT.out.rds                  //    path: *.rds
 
     versions                      = ch_versions                                   // channel: [ versions.yml ]
 }
diff --git a/workflows/rnaseq.nf b/workflows/rnaseq.nf
@@ -205,6 +205,20 @@ workflow RNASEQ {
     .set { ch_cat_fastq }
     ch_versions = ch_versions.mix(CAT_FASTQ.out.versions.first().ifEmpty(null))
 
+
+    // Add sub-workflow here that sub-samples FastQ files and runs SALMON_QUANT
+    // Filter channels beforehand for anything with strandedness not set to 'auto'
+    // QUANTIFY_SALMON
+    //     .out
+    //     .json_info
+    //     //.filter { meta, json -> meta.strandedness != 'auto' }  // Add this filter right at the beginning
+    //     .filter { meta, json -> meta.strandedness != 'auto' }
+    //     .map { meta, json ->
+    //         meta.strandedness = WorkflowRnaseq.getSalmonInferredStrandedness(json)
+    //         return [ meta, json ]
+    //     }
+    //     .view()
+
     //
     // SUBWORKFLOW: Read QC, extract UMI and trim adapters
     //