Merge pull request #708 from dialvarezs/dev-exclude-unbinned

Add parameter to exclude unbinned data from post-binning steps
nf-core · Oct 30, 2024 · cd5ebae · cd5ebae
2 parents f0ca539 + 1a00c14
commit cd5ebae
Show file tree

Hide file tree

Showing 4 changed files with 24 additions and 9 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -7,10 +7,14 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### `Added`
 
+- [#708](https://github.com/nf-core/mag/pull/708) - Added `--exclude_unbins_from_postbinning` parameter to exclude unbinned contigs from post-binning processes, speeding up Prokka in some cases (added by @dialvarezs)
+
 ### `Changed`
 
 ### `Fixed`
 
+- [#708](https://github.com/nf-core/mag/pull/708) - Fixed channel passed as GUNC input (added by @dialvarezs)
+
 ### `Dependencies`
 
 ### `Deprecated`

diff --git a/nextflow.config b/nextflow.config
@@ -121,6 +121,7 @@ params {
     refine_bins_dastool             = false
     refine_bins_dastool_threshold   = 0.5
     postbinning_input               = 'raw_bins_only'
+    exclude_unbins_from_postbinning   = false
 
     // Bin QC
     skip_binqc                           = false

diff --git a/nextflow_schema.json b/nextflow_schema.json
@@ -704,6 +704,12 @@
                     "type": "integer",
                     "default": 3000,
                     "description": "Minimum contig length for Tiara to use for domain classification. For accurate classification, should be longer than 3000 bp."
+                },
+                "exclude_unbins_from_postbinning": {
+                    "type": "boolean",
+                    "description": "Exclude unbinned contigs in the post-binning steps (bin QC, taxonomic classification, and annotation steps).",
+                    "help": "If you're not interested in assemby results that are not considered 'genome level', excluding unbinned contigs can greatly speed up downstream steps such as Prokka, that can be quite slow and spin up many tasks.",
+                    "default": false
                 }
             }
         },

diff --git a/workflows/mag.nf b/workflows/mag.nf
@@ -769,15 +769,19 @@ workflow MAG {
             ch_input_for_postbinning_bins_unbins = ch_binning_results_bins.mix(ch_binning_results_unbins)
         }
 
-        DEPTHS(ch_input_for_postbinning_bins_unbins, BINNING.out.metabat2depths, ch_short_reads)
+        ch_input_for_postbinning = params.exclude_unbins_from_postbinning
+            ? ch_input_for_postbinning_bins
+            : ch_input_for_postbinning_bins_unbins
+
+        DEPTHS(ch_input_for_postbinning, BINNING.out.metabat2depths, ch_short_reads)
         ch_input_for_binsummary = DEPTHS.out.depths_summary
         ch_versions = ch_versions.mix(DEPTHS.out.versions)
 
         /*
         * Bin QC subworkflows: for checking bin completeness with either BUSCO, CHECKM, and/or GUNC
         */
 
-        ch_input_bins_for_qc = ch_input_for_postbinning_bins_unbins.transpose()
+        ch_input_bins_for_qc = ch_input_for_postbinning.transpose()
 
         if (!params.skip_binqc && params.binqc_tool == 'busco') {
             /*
@@ -821,16 +825,16 @@ workflow MAG {
             ch_versions = ch_versions.mix(GUNC_QC.out.versions)
         }
         else if (params.run_gunc) {
-            ch_input_bins_for_gunc = ch_input_for_postbinning_bins_unbins.filter { meta, bins ->
+            ch_input_bins_for_gunc = ch_input_for_postbinning.filter { meta, bins ->
                 meta.domain != "eukarya"
             }
-            GUNC_QC(ch_input_bins_for_qc, ch_gunc_db, [])
+            GUNC_QC(ch_input_bins_for_gunc, ch_gunc_db, [])
             ch_versions = ch_versions.mix(GUNC_QC.out.versions)
         }
 
         ch_quast_bins_summary = Channel.empty()
         if (!params.skip_quast) {
-            ch_input_for_quast_bins = ch_input_for_postbinning_bins_unbins
+            ch_input_for_quast_bins = ch_input_for_postbinning
                 .groupTuple()
                 .map { meta, bins ->
                     def new_bins = bins.flatten()
@@ -859,7 +863,7 @@ workflow MAG {
             ch_cat_db = CAT_DB_GENERATE.out.db
         }
         CAT(
-            ch_input_for_postbinning_bins_unbins,
+            ch_input_for_postbinning,
             ch_cat_db
         )
         // Group all classification results for each sample in a single file
@@ -890,7 +894,7 @@ workflow MAG {
             ch_gtdbtk_summary = Channel.empty()
             if (gtdb) {
 
-                ch_gtdb_bins = ch_input_for_postbinning_bins_unbins.filter { meta, bins ->
+                ch_gtdb_bins = ch_input_for_postbinning.filter { meta, bins ->
                     meta.domain != "eukarya"
                 }
 
@@ -925,7 +929,7 @@ workflow MAG {
          */
 
         if (!params.skip_prokka) {
-            ch_bins_for_prokka = ch_input_for_postbinning_bins_unbins
+            ch_bins_for_prokka = ch_input_for_postbinning
                 .transpose()
                 .map { meta, bin ->
                     def meta_new = meta + [id: bin.getBaseName()]
@@ -944,7 +948,7 @@ workflow MAG {
         }
 
         if (!params.skip_metaeuk && (params.metaeuk_db || params.metaeuk_mmseqs_db)) {
-            ch_bins_for_metaeuk = ch_input_for_postbinning_bins_unbins
+            ch_bins_for_metaeuk = ch_input_for_postbinning
                 .transpose()
                 .filter { meta, bin ->
                     meta.domain in ["eukarya", "unclassified"]