Integrate GUNC in BIN_QC subworkflow

nf-core · Nov 1, 2024 · 0eb167a · 0eb167a
1 parent 3702329
commit 0eb167a
Show file tree

Hide file tree

Showing 5 changed files with 99 additions and 111 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -7,17 +7,22 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### `Added`
 
+- [#707](https://github.com/nf-core/mag/pull/707) - Make Bin QC a subworkflow (added by @dialvarezs)
 - [#707](https://github.com/nf-core/mag/pull/707) - Added CheckM2 as an alternative bin completeness and QC tool (added by @dialvarezs)
 - [#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)
+- [#707](https://github.com/nf-core/mag/pull/708) - Fixed channel passed as GUNC input (added by @dialvarezs)
 
 ### `Dependencies`
 
+| Tool    | Previous version | New version |
+| ------- | ---------------- | ----------- |
+| CheckM2 |                  | 1.0.2       |
+
 ### `Deprecated`
 
 ## 3.2.1 [2024-10-30]
@@ -32,10 +37,6 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### `Dependencies`
 
-| Tool    | Previous version | New version |
-| ------- | ---------------- | ----------- |
-| CheckM2 |                  | 1.0.2       |
-
 ### `Deprecated`
 
 ## 3.2.0 [2024-10-27]

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

diff --git a/subworkflows/local/bin_qc.nf b/subworkflows/local/bin_qc.nf
@@ -1,5 +1,5 @@
 /*
- * BUSCO/CheckM/CheckM2: Quantitative measures for the assessment of genome assembly
+ * BUSCO/CheckM/CheckM2/GUNC: Quantitative measures for the assessment of genome assembly
  */
 
 include { BUSCO_DB_PREPARATION              } from '../../modules/local/busco_db_preparation'
@@ -10,33 +10,42 @@ include { CHECKM_QA                         } from '../../modules/nf-core/checkm
 include { CHECKM_LINEAGEWF                  } from '../../modules/nf-core/checkm/lineagewf/main'
 include { CHECKM2_PREDICT                   } from '../../modules/nf-core/checkm2/predict/main'
 include { COMBINE_TSV as COMBINE_BINQC_TSV  } from '../../modules/local/combine_tsv'
+include { GUNC_DOWNLOADDB                   } from '../../modules/nf-core/gunc/downloaddb/main'
+include { GUNC_RUN                          } from '../../modules/nf-core/gunc/run/main'
+include { GUNC_MERGECHECKM                  } from '../../modules/nf-core/gunc/mergecheckm/main'
+
 
 workflow BIN_QC {
     take:
-    bins       // channel: [ val(meta), path(bin) ]
-    checkm_db
-    checkm2_db
-    busco_db
+    ch_bins        // [ [ meta] , fasta           ], input bins (mandatory)
+    ch_checkm_db   // [ db                        ], presupplied CheckM database (optional)
+    ch_checkm2_db  // [ [meta]  , db              ], presupplied CheckM2 database (optional)
+    ch_busco_db    // [ [meta]  , db              ], presupplied BUSCO database (optional)
+    ch_gunc_db     // [ db                        ], presupplied GUNC database (optional)
 
     main:
+    qc_summary = []
+    ch_input_bins_for_qc = ch_bins.transpose()
     ch_versions = Channel.empty()
-    multiqc_reports = []
-    checkm_tsv = []
+    ch_multiqc_files = Channel.empty()
+
 
     if (params.binqc_tool == "busco") {
-        // BUSCO workflow
-        if (!busco_db.isEmpty()) {
-            if (busco_db.extension in ['gz', 'tgz']) {
+        /*
+         * BUSCO
+         */
+        if (!ch_busco_db.isEmpty()) {
+            if (ch_busco_db.extension in ['gz', 'tgz']) {
                 // Expects to be tar.gz!
-                BUSCO_DB_PREPARATION(busco_db)
+                BUSCO_DB_PREPARATION(ch_busco_db)
                 ch_db_for_busco = BUSCO_DB_PREPARATION.out.db.map { meta, db ->
                     [[id: meta, lineage: 'Y'], db]
                 }
             }
-            else if (busco_db.isDirectory()) {
+            else if (ch_busco_db.isDirectory()) {
                 // Set meta to match expected channel cardinality for BUSCO
                 ch_db_for_busco = Channel
-                    .of(busco_db)
+                    .of(ch_busco_db)
                     .map { db ->
                         def basename = db.getBaseName()
                         def lineage = basename.contains('odb10') ? 'Y' : 'N'
@@ -48,8 +57,7 @@ workflow BIN_QC {
         else {
             // Set BUSCO database to empty to allow for --auto-lineage
             ch_db_for_busco = Channel
-                .of([])
-                .map { _empty_db -> [[lineage: ''], []] }
+                .of([[lineage: ''], []])
                 .collect()
         }
 
@@ -63,21 +71,25 @@ workflow BIN_QC {
             BUSCO_SAVE_DOWNLOAD(ch_downloads)
         }
 
-        BUSCO(bins, ch_db_for_busco)
+        BUSCO(ch_input_bins_for_qc, ch_db_for_busco)
 
         BUSCO_SUMMARY(
             BUSCO.out.summary_domain.map { it[1] }.collect().ifEmpty([]),
             BUSCO.out.summary_specific.map { it[1] }.collect().ifEmpty([]),
             BUSCO.out.failed_bin.map { it[1] }.collect().ifEmpty([])
         )
 
-        multiqc_reports = BUSCO.out.summary_domain.mix(BUSCO.out.summary_specific).map{ it[1] }
-        summary  = BUSCO_SUMMARY.out.summary
+        ch_multiqc_files = ch_multiqc_files.mix(
+            BUSCO.out.summary_domain.mix(BUSCO.out.summary_specific).map{ it[1] }
+        )
+        qc_summary = BUSCO_SUMMARY.out.summary
         ch_versions = ch_versions.mix(BUSCO.out.versions.first())
     }
     else if (params.binqc_tool == "checkm") {
-        // CheckM workflow
-        ch_bins_for_checkmlineagewf = bins
+        /*
+         * CheckM
+         */
+        ch_bins_for_checkmlineagewf = ch_input_bins_for_qc
             .groupTuple()
             .filter { meta, _bins ->
                     meta.domain != "eukarya"
@@ -87,7 +99,7 @@ workflow BIN_QC {
                 ext: fa.extension.unique().join("")  // the pipeline ensures that all bins will have the same extension
             }
 
-        CHECKM_LINEAGEWF(ch_bins_for_checkmlineagewf.reads, ch_bins_for_checkmlineagewf.ext, checkm_db)
+        CHECKM_LINEAGEWF(ch_bins_for_checkmlineagewf.reads, ch_bins_for_checkmlineagewf.ext, ch_checkm_db)
         ch_versions = ch_versions.mix(CHECKM_LINEAGEWF.out.versions.first())
 
         ch_checkmqa_input = CHECKM_LINEAGEWF.out.checkm_output
@@ -100,23 +112,65 @@ workflow BIN_QC {
 
         COMBINE_BINQC_TSV(CHECKM_QA.out.output.map { it[1] }.collect())
 
-        summary = COMBINE_BINQC_TSV.out.combined
+        qc_summary = COMBINE_BINQC_TSV.out.combined
         ch_versions = ch_versions.mix(CHECKM_QA.out.versions.first())
-        checkm_tsv = CHECKM_QA.out.output
     }
     else if (params.binqc_tool == "checkm2") {
-        // CheckM2 workflow
-        CHECKM2_PREDICT(bins.groupTuple(), checkm2_db)
+        /*
+         * CheckM2
+         */
+        CHECKM2_PREDICT(ch_input_bins_for_qc.groupTuple(), ch_checkm2_db)
 
         COMBINE_BINQC_TSV(CHECKM2_PREDICT.out.checkm2_tsv.map { it[1] }.collect())
 
-        summary = COMBINE_BINQC_TSV.out.combined
+        qc_summary = COMBINE_BINQC_TSV.out.combined
         ch_versions = ch_versions.mix(CHECKM2_PREDICT.out.versions.first())
     }
 
+    if (params.run_gunc) {
+        /*
+         * GUNC
+         */
+        ch_input_bins_for_gunc = ch_bins
+            .filter { meta, _bins ->
+                meta.domain != "eukarya"
+            }
+            .flatMap { meta, bins ->
+                bins.collect { bin -> [meta, bin] }
+            }
+
+        if ( params.gunc_db ) {
+            ch_db_for_gunc = ch_gunc_db
+        }
+        else {
+            ch_db_for_gunc = GUNC_DOWNLOADDB(params.gunc_database_type).db
+            ch_versions.mix(GUNC_DOWNLOADDB.out.versions)
+        }
+
+        GUNC_RUN(ch_input_bins_for_gunc, ch_db_for_gunc)
+        ch_versions.mix(GUNC_RUN.out.versions)
+
+        // Make sure to keep directory in sync with modules.conf
+        GUNC_RUN.out.maxcss_level_tsv
+            .map{it[1]}
+            .collectFile(name: "gunc_summary.tsv", keepHeader: true, storeDir: "${params.outdir}/GenomeBinning/QC/")
+
+        if ( params.binqc_tool == 'checkm' ) {
+            ch_input_to_mergecheckm = GUNC_RUN.out.maxcss_level_tsv.combine(CHECKM_QA.out.output, by: 0)
+
+            GUNC_MERGECHECKM(ch_input_to_mergecheckm)
+            ch_versions.mix(GUNC_MERGECHECKM.out.versions)
+
+            // Make sure to keep directory in sync with modules.conf
+            GUNC_MERGECHECKM.out.tsv
+                .map{it[1]}
+                .collectFile(name: "gunc_checkm_summary.tsv", keepHeader: true, storeDir: "${params.outdir}/GenomeBinning/QC/")
+        }
+    }
+
+
     emit:
-    summary     = summary
-    checkm_tsv  = checkm_tsv
-    multiqc     = multiqc_reports
-    versions    = ch_versions
+    qc_summary      = qc_summary
+    multiqc_files   = ch_multiqc_files
+    versions        = ch_versions
 }
diff --git a/subworkflows/local/gunc_qc.nf b/subworkflows/local/gunc_qc.nf
diff --git a/workflows/mag.nf b/workflows/mag.nf
@@ -17,7 +17,6 @@ include { BINNING                                               } from '../subwo
 include { BIN_QC                                                } from '../subworkflows/local/bin_qc'
 include { BINNING_REFINEMENT                                    } from '../subworkflows/local/binning_refinement'
 include { VIRUS_IDENTIFICATION                                  } from '../subworkflows/local/virus_identification'
-include { GUNC_QC                                               } from '../subworkflows/local/gunc_qc'
 include { GTDBTK                                                } from '../subworkflows/local/gtdbtk'
 include { ANCIENT_DNA_ASSEMBLY_VALIDATION                       } from '../subworkflows/local/ancient_dna'
 include { DOMAIN_CLASSIFICATION                                 } from '../subworkflows/local/domain_classification'
@@ -649,7 +648,7 @@ workflow MAG {
     ================================================================================
     */
 
-    bin_qc_summary = Channel.empty()
+    ch_bin_qc_summary = Channel.empty()
 
     if (!params.skip_binning || params.ancient_dna) {
         BINNING_PREPARATION(
@@ -792,32 +791,17 @@ workflow MAG {
         * Bin QC subworkflows: for checking bin completeness with either BUSCO, CHECKM, CHECKM2, and/or GUNC
         */
 
-        ch_input_bins_for_qc = ch_input_for_postbinning.transpose()
-
         BIN_QC(
-            ch_input_bins_for_qc,
+            ch_input_for_postbinning,
             ch_checkm_db,
             ch_checkm2_db,
-            ch_busco_db
+            ch_busco_db,
+            ch_gunc_db
         )
 
-        bin_qc_summary = BIN_QC.out.summary
+        ch_bin_qc_summary = BIN_QC.out.qc_summary
         ch_versions = ch_versions.mix(BIN_QC.out.versions)
 
-        if (params.run_gunc) {
-            ch_input_bins_for_gunc = ch_input_for_postbinning.filter { meta, _bins ->
-                meta.domain != "eukarya"
-            }
-
-            GUNC_QC(
-                ch_input_bins_for_gunc,
-                ch_gunc_db,
-                params.binqc_tool == 'checkm' ? BIN_QC.out.checkm_tsv : []
-            )
-
-            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
@@ -886,7 +870,7 @@ workflow MAG {
 
                 GTDBTK(
                     ch_gtdb_bins,
-                    bin_qc_summary,
+                    ch_bin_qc_summary,
                     gtdb,
                     gtdb_mash
                 )
@@ -901,7 +885,7 @@ workflow MAG {
         if ((!params.skip_binqc) || !params.skip_quast || !params.skip_gtdbtk) {
             BIN_SUMMARY(
                 ch_input_for_binsummary,
-                bin_qc_summary.ifEmpty([]),
+                ch_bin_qc_summary.ifEmpty([]),
                 ch_quast_bins_summary.ifEmpty([]),
                 ch_gtdbtk_summary.ifEmpty([]),
                 ch_cat_global_summary.ifEmpty([])
@@ -1043,7 +1027,7 @@ workflow MAG {
     }
 
     if (!params.skip_binning && !params.skip_binqc && params.binqc_tool == 'busco') {
-        ch_multiqc_files = ch_multiqc_files.mix(BIN_QC.out.multiqc.collect().ifEmpty([]))
+        ch_multiqc_files = ch_multiqc_files.mix(BIN_QC.out.multiqc_files.collect().ifEmpty([]))
     }