[Fix gprofiler] Empty output table is now always added (nf-core#4583)

* Bugfix, empty output table is now always added * Replaced multiple contrast params with one prefix param
luisas · Dec 13, 2023 · ef6053b · ef6053b
1 parent 4be1afb
commit ef6053b
Show file tree

Hide file tree

Showing 4 changed files with 38 additions and 50 deletions.
diff --git a/modules/nf-core/gprofiler2/gost/main.nf b/modules/nf-core/gprofiler2/gost/main.nf
@@ -1,5 +1,5 @@
 process GPROFILER2_GOST {
-    tag "$meta"
+    tag "$meta.id"
     label 'process_single'
 
     conda "${moduleDir}/environment.yml"
@@ -8,21 +8,20 @@ process GPROFILER2_GOST {
         'biocontainers/mulled-v2-3712554873398d849d0d11b22440f41febbc4ede:aa19bb8afc0ec6456a4f3cd650f7577c3bbdd4f3-0' }"
 
     input:
-    tuple val(meta), val(contrast_variable), val(reference), val(target)
-    tuple val(meta1), path(de_file)
+    tuple val(meta), path(de_file)
     path(gmt_file)
     path(background_file)
 
     output:
-    tuple val(meta), path("*gprofiler2.*all_enriched_pathways.tsv") , emit: all_enrich
-    tuple val(meta), path("*gprofiler2.*gost_results.rds")          , emit: rds         , optional: true
-    tuple val(meta), path("*gprofiler2.*gostplot.png")              , emit: plot_png    , optional: true
-    tuple val(meta), path("*gprofiler2.*gostplot.html")             , emit: plot_html   , optional: true
-    tuple val(meta), path("*gprofiler2.*sub_enriched_pathways.tsv") , emit: sub_enrich  , optional: true
-    tuple val(meta), path("*gprofiler2.*sub_enriched_pathways.png") , emit: sub_plot    , optional: true
-    tuple val(meta), path("*ENSG_filtered.gmt")                     , emit: filtered_gmt, optional: true
-    tuple val(meta), path("*R_sessionInfo.log")                     , emit: session_info
-    path "versions.yml"                                             , emit: versions
+    tuple val(meta), path("*.gprofiler2.all_enriched_pathways.tsv")     , emit: all_enrich
+    tuple val(meta), path("*.gprofiler2.gost_results.rds")              , emit: rds         , optional: true
+    tuple val(meta), path("*.gprofiler2.gostplot.png")                  , emit: plot_png    , optional: true
+    tuple val(meta), path("*.gprofiler2.gostplot.html")                 , emit: plot_html   , optional: true
+    tuple val(meta), path("*.gprofiler2.*.sub_enriched_pathways.tsv")   , emit: sub_enrich  , optional: true
+    tuple val(meta), path("*.gprofiler2.*.sub_enriched_pathways.png")   , emit: sub_plot    , optional: true
+    tuple val(meta), path("*ENSG_filtered.gmt")                         , emit: filtered_gmt, optional: true
+    tuple val(meta), path("*R_sessionInfo.log")                         , emit: session_info
+    path "versions.yml"                                                 , emit: versions
 
     when:
     task.ext.when == null || task.ext.when

diff --git a/modules/nf-core/gprofiler2/gost/templates/gprofiler2_gost.R b/modules/nf-core/gprofiler2/gost/templates/gprofiler2_gost.R
@@ -126,10 +126,7 @@ opt <- list(
     de_id_column = 'gene_id',
     organism = NULL,
     sources = NULL,
-    contrast_variable = '$contrast_variable',
-    reference_level = '$reference',
-    target_level = '$target',
-    blocking_variables = NULL,
+    output_prefix = ifelse('$task.ext.prefix' == 'null', '$meta.id', '$task.ext.prefix'),
     significant = T,
     measure_underrepresentation = F,
     correction_method = 'gSCS',
@@ -162,9 +159,8 @@ for ( ao in names(args_opt)) {
         opt[[ao]] <- args_opt[[ao]]
     }
 }
-
 # Check if required parameters have been provided
-required_opts <- c('contrast_variable', 'reference_level', 'target_level')
+required_opts <- c('output_prefix')
 missing <- required_opts[unlist(lapply(opt[required_opts], is.null)) | ! required_opts %in% names(opt)]
 
 if (length(missing) > 0) {
@@ -206,13 +202,10 @@ de.genes <-
         file = opt\$de_file
     )
 
-# Create prefix from contrast var, reference and target as well as blocking (if provided)
-contrast_name <- paste(opt\$contrast_variable, opt\$reference_level, opt\$target_level, sep = '_')
-output_prefix <- paste('gprofiler2', contrast_name, sep = '.')
-if (!is.null(opt\$blocking_variables)) {
-    blocking_variables <- paste(make.names(unlist(strsplit(opt\$blocking_variables, split = ','))), collapse = '_')
-    output_prefix <- paste(output_prefix, blocking_variables, sep= '_')
-}
+output_prefix <- paste0(opt\$output_prefix, ".gprofiler2")
+
+# Create empty output table as it is a mandatory output
+file.create(paste(output_prefix, 'all_enriched_pathways', 'tsv', sep = '.'))
 
 if (nrow(de.genes) > 0) {
 
@@ -328,8 +321,7 @@ if (nrow(de.genes) > 0) {
 
     # Name the query as it will otherwise be called 'query_1' which will also determine the gostplot title
     q <- list(query)
-    names(q) <- c(paste0(contrast_name, ifelse(!is.null(opt\$blocking_variables), paste0("_", blocking_variables), "")))
-
+    names(q) <- c(output_prefix)
     gost_results <- gost(
         query=q,
         organism=token,
@@ -365,7 +357,7 @@ if (nrow(de.genes) > 0) {
 
         # R object for other processes to use
 
-        saveRDS(gost_results, file = paste(opt\$contrast_variable, 'gprofiler2.gost_results.rds', sep = '.'))
+        saveRDS(gost_results, file = paste(output_prefix, 'gost_results.rds', sep = '.'))
 
         # Write full enrichment table (except parents column as that one throws an error)
 
@@ -385,6 +377,7 @@ if (nrow(de.genes) > 0) {
 
         # Iterate over the enrichment results by source and save separate tables
         for (df in split(gost_results\$result, gost_results\$result\$source)){
+
             db_source <- df\$source[1]
             df_subset <- data.frame(
                 Pathway_name = df\$term_name,
@@ -430,8 +423,6 @@ if (nrow(de.genes) > 0) {
         }
     }
 } else {
-    # Create empty output table as it is a mandatory output
-    file.create(paste(output_prefix, 'all_enriched_pathways', 'tsv', sep = '.'))
     print("No differential features found, pathway enrichment analysis with gprofiler2 will be skipped.")
 }
 

diff --git a/tests/modules/nf-core/gprofiler2/gost/main.nf b/tests/modules/nf-core/gprofiler2/gost/main.nf
@@ -7,12 +7,11 @@ include { GPROFILER2_GOST } from '../../../../../modules/nf-core/gprofiler2/gost
 workflow test_gprofiler2_gost {
     contrasts = [ [ id:'test', reference:'r', target:'t' ], 'test', 'r', 't' ]
     input = [
-        [ id:'test', reference:'r', target:'t' ], // meta map
+        [ id:'test_r_t', reference:'r', target:'t' ], // meta map
         file(params.test_data['mus_musculus']['genome']['deseq_results'], checkIfExists: true)
     ]
 
     GPROFILER2_GOST (
-        contrasts,
         input,
         [],
         []
@@ -22,13 +21,12 @@ workflow test_gprofiler2_gost {
 workflow test_gprofiler2_gost_backgroundmatrix {
     contrasts = [ [ id:'test', reference:'r', target:'t' ], 'test', 'r', 't' ]
     input = [
-        [ id:'test', reference:'r', target:'t' ], // meta map
+        [ id:'test_r_t', reference:'r', target:'t' ], // meta map
         file(params.test_data['mus_musculus']['genome']['deseq_results'], checkIfExists: true)
     ]
     ch_background = Channel.from(file(params.test_data['mus_musculus']['genome']['rnaseq_matrix'], checkIfExists: true))
 
     GPROFILER2_GOST (
-        contrasts,
         input,
         [],
         ch_background

diff --git a/tests/modules/nf-core/gprofiler2/gost/test.yml b/tests/modules/nf-core/gprofiler2/gost/test.yml
@@ -6,25 +6,25 @@
   files:
     - path: output/gprofiler2/R_sessionInfo.log
       contains: ["ggplot2_3.4.3", "gprofiler2_0.2.2"]
-    - path: output/gprofiler2/gprofiler2.test_r_t.CORUM.sub_enriched_pathways.png
-    - path: output/gprofiler2/gprofiler2.test_r_t.CORUM.sub_enriched_pathways.tsv
+    - path: output/gprofiler2/test_r_t.gprofiler2.CORUM.sub_enriched_pathways.png
+    - path: output/gprofiler2/test_r_t.gprofiler2.CORUM.sub_enriched_pathways.tsv
       md5sum: b0619e5e1424ac18f6cb52dac87248aa
-    - path: output/gprofiler2/gprofiler2.test_r_t.REAC.sub_enriched_pathways.png
-    - path: output/gprofiler2/gprofiler2.test_r_t.REAC.sub_enriched_pathways.tsv
+    - path: output/gprofiler2/test_r_t.gprofiler2.REAC.sub_enriched_pathways.png
+    - path: output/gprofiler2/test_r_t.gprofiler2.REAC.sub_enriched_pathways.tsv
       md5sum: 0e4f2887e74e00fe8d0ad93771342f4b
-    - path: output/gprofiler2/gprofiler2.test_r_t.all_enriched_pathways.tsv
-      md5sum: 231bdf9e12394dbb37d9bd427da36fb5
-    - path: output/gprofiler2/gprofiler2.test_r_t.gostplot.html
+    - path: output/gprofiler2/test_r_t.gprofiler2.all_enriched_pathways.tsv
+      md5sum: bbfa48be9e30e9898ecf63709bb99331
+    - path: output/gprofiler2/test_r_t.gprofiler2.gostplot.html
       contains:
         [
           "Neurotransmitter receptors and postsynaptic signal transmission",
           "The phototransduction cascade",
           "Muscle contraction",
         ]
-    - path: output/gprofiler2/gprofiler2.test_r_t.gostplot.png
+    - path: output/gprofiler2/test_r_t.gprofiler2.gostplot.png
     - path: output/gprofiler2/gprofiler_full_mmusculus.CORUM_REAC.ENSG_filtered.gmt
       md5sum: 455f9b94af175e78cc551cf7f79c3203
-    - path: output/gprofiler2/test.gprofiler2.gost_results.rds
+    - path: output/gprofiler2/test_r_t.gprofiler2.gost_results.rds
     - path: output/gprofiler2/versions.yml
 
 - name: gprofiler2 gost test_gprofiler2_gost_backgroundmatrix
@@ -35,13 +35,13 @@
   files:
     - path: output/gprofiler2/R_sessionInfo.log
       contains: ["ggplot2_3.4.3", "gprofiler2_0.2.2"]
-    - path: output/gprofiler2/gprofiler2.test_r_t.KEGG.sub_enriched_pathways.png
-    - path: output/gprofiler2/gprofiler2.test_r_t.KEGG.sub_enriched_pathways.tsv
+    - path: output/gprofiler2/test_r_t.gprofiler2.KEGG.sub_enriched_pathways.png
+    - path: output/gprofiler2/test_r_t.gprofiler2.KEGG.sub_enriched_pathways.tsv
       md5sum: fa7c0457da981ca00b209fadafe419bd
-    - path: output/gprofiler2/gprofiler2.test_r_t.all_enriched_pathways.tsv
-      md5sum: 6eeff6d8067f161425184039d68d5d1f
-    - path: output/gprofiler2/gprofiler2.test_r_t.gostplot.html # This plot is empty, will not add contains
-    - path: output/gprofiler2/gprofiler2.test_r_t.gostplot.png
+    - path: output/gprofiler2/test_r_t.gprofiler2.all_enriched_pathways.tsv
+      md5sum: fd611ba563ed8ec2c27dbf8d99b869dd
+    - path: output/gprofiler2/test_r_t.gprofiler2.gostplot.html # This plot is empty, will not add contains
+    - path: output/gprofiler2/test_r_t.gprofiler2.gostplot.png
     - path: output/gprofiler2/gprofiler_full_mmusculus.KEGG.ENSG_filtered.gmt
-    - path: output/gprofiler2/test.gprofiler2.gost_results.rds
+    - path: output/gprofiler2/test_r_t.gprofiler2.gost_results.rds
     - path: output/gprofiler2/versions.yml