Several busco-related fixes

nf-core · Dec 12, 2024 · 25b72f1 · 25b72f1
1 parent f85d7b0
commit 25b72f1
Show file tree

Hide file tree

Showing 10 changed files with 23 additions and 52 deletions.
diff --git a/bin/combine_tables.py b/bin/combine_tables.py
@@ -120,11 +120,11 @@ def main(args=None):
     if args.busco_summary:
         busco_results = pd.read_csv(args.busco_summary, sep="\t")
         if not bins.equals(
-            busco_results["GenomeBin"].sort_values().reset_index(drop=True)
+            busco_results["Input_file"].sort_values().reset_index(drop=True)
         ):
             sys.exit("Bins in BUSCO summary do not match bins in bin depths summary!")
         results = pd.merge(
-            results, busco_results, left_on="bin", right_on="GenomeBin", how="outer"
+            results, busco_results, left_on="bin", right_on="Input_file", how="outer"
         )  # assuming depths for all bins are given
 
     if args.checkm_summary:

diff --git a/conf/base.config b/conf/base.config
@@ -157,9 +157,8 @@ process {
         memory = { 16.GB * task.attempt }
     }
     withName: BUSCO {
-        cpus          = { 4 * task.attempt }
+        cpus          = { 8 * task.attempt }
         memory        = { 10.GB * task.attempt }
-        errorStrategy = { task.exitStatus in ((130..145) + 104 + 250) ? 'retry' : 'ignore' }
     }
 
     withName: MAXBIN2 {

diff --git a/conf/modules.config b/conf/modules.config
@@ -375,10 +375,6 @@ process {
         ]
     }
 
-    withName: BUSCO_DB_PREPARATION {
-        publishDir = [path: { "${params.outdir}/GenomeBinning/QC/BUSCO" }, mode: params.publish_dir_mode, pattern: "*.tar.gz"]
-    }
-
     withName: BUSCO {
         ext.args   = [
             params.busco_db ? '--offline' : ''

diff --git a/modules.json b/modules.json
@@ -38,7 +38,7 @@
                     },
                     "busco/busco": {
                         "branch": "master",
-                        "git_sha": "abd32eb8341f565fed86738307a39404fc1da0ce",
+                        "git_sha": "9a19690b0a3fae05fa1a6ad90a0720d681429e31",
                         "installed_by": ["modules"]
                     },
                     "cat/fastq": {

diff --git a/modules/local/combine_tsv.nf b/modules/local/combine_tsv.nf
@@ -7,7 +7,7 @@ process COMBINE_TSV {
         'biocontainers/bioawk:1.0--hed695b0_5' }"
 
     input:
-    path(bin_summaries)
+    path(bin_summaries, stageAs: "bin_summaries/*.tsv")
 
     output:
     path("*.tsv")      , emit: combined

diff --git a/modules/nf-core/busco/busco/main.nf b/modules/nf-core/busco/busco/main.nf
diff --git a/nextflow.config b/nextflow.config
@@ -132,6 +132,7 @@ params {
     busco_auto_lineage_prok              = false
     save_busco_db                        = false
     busco_clean                          = false
+    busco_lineage                        = 'auto'
     checkm_download_url                  = "https://zenodo.org/records/7401545/files/checkm_data_2015_01_16.tar.gz"
     checkm_db                            = null
     save_checkm_data                     = false

diff --git a/nextflow_schema.json b/nextflow_schema.json
@@ -744,6 +744,11 @@
                     "description": "Download URL for BUSCO lineage dataset, or path to a tar.gz archive, or local directory containing already downloaded and unpacked lineage datasets.",
                     "help_text": "E.g. https://busco-data.ezlab.org/v5/data/lineages/bacteria_odb10.2024-01-08.tar.gz or '/path/to/buscodb' (files still need to be unpacked manually). Available databases are listed here: https://busco-data.ezlab.org/v5/data/lineages/."
                 },
+                "busco_lineage": {
+                    "type": "string",
+                    "default": "auto",
+                    "description": "Lineage to use for BUSCO (for example, bacteria_odb12). Additionally, supports 'auto' and 'auto_prok' for automated lineage selection."
+                },
                 "busco_auto_lineage_prok": {
                     "type": "boolean",
                     "description": "Run BUSCO with automated lineage selection, but ignoring eukaryotes (saves runtime)."

diff --git a/subworkflows/local/busco_qc.nf b/subworkflows/local/busco_qc.nf
@@ -16,49 +16,19 @@ workflow BUSCO_QC {
     main:
     if (!busco_db.isEmpty()) {
         if (busco_db.extension in ['gz', 'tgz']) {
-            BUSCO_UNTAR(busco_db.map { db -> [[id: 'busco_db'], db] })
+            BUSCO_UNTAR([[id: 'busco_db'], busco_db])
 
-            busco_lineage = busco_db.getSimpleName()
+            if (busco_db.getSimpleName().contains('odb')) {
+                busco_lineage = busco_db.getSimpleName()
+            }
             busco_db = BUSCO_UNTAR.out.untar.map { it[1] }
-
-            // Expects to be tar.gz!
-            // ch_db_for_busco = BUSCO_DB_PREPARATION(busco_db).db.map { meta, db ->
-            //     def meta_new = [:]
-            //     meta_new['id'] = meta
-            //     meta_new['lineage'] = 'Y'
-            //     [meta_new, db]
-            // }
         }
-        // else if (busco_db.isDirectory()) {
-        //     // Set meta to match expected channel cardinality for BUSCO
-        //     ch_db_for_busco = Channel
-        //         .of(busco_db)
-        //         .map { db ->
-        //             def meta = [:]
-        //             meta['id'] = db.getBaseName()
-        //             if (meta['id'].contains('odb10') == true) {
-        //                 meta['lineage'] = 'Y'
-        //             }
-        //             else {
-        //                 meta['lineage'] = 'N'
-        //             }
-        //             [meta, db]
-        //         }
-        //         .collect()
-        // }
-        // BUSCO_BUSCO(bins, params.busco_lineage)
+        else if (busco_db.isDirectory()) {
+            if (busco_db.name.contains('odb')) {
+                busco_lineage = busco_db.name
+            }
+        }
     }
-    // else {
-        // Set BUSCO database to empty to allow for --auto-lineage
-    //     ch_db_for_busco = Channel
-    //         .of([])
-    //         .map { empty_db ->
-    //             def meta = [:]
-    //             meta['lineage'] = ''
-    //             [meta, []]
-    //         }
-    //         .collect()
-    // }
 
     BUSCO_BUSCO( bins, 'genome', busco_lineage, busco_db, [] )
 
@@ -68,7 +38,7 @@ workflow BUSCO_QC {
     //     BUSCO_SAVE_DOWNLOAD(ch_downloads)
     // }
 
-    COMBINE_BUSCO_TSV(BUSCO_BUSCO.out.batch_summary.collect())
+    COMBINE_BUSCO_TSV(BUSCO_BUSCO.out.batch_summary.map { it[1] }.collect())
 
     emit:
     summary    = COMBINE_BUSCO_TSV.out.combined

diff --git a/workflows/mag.nf b/workflows/mag.nf
@@ -654,7 +654,7 @@ workflow MAG {
             BUSCO_QC(
                 ch_input_bins_for_qc.groupTuple(),
                 ch_busco_db,
-                'auto_prok'
+                params.busco_lineage
             )
             ch_busco_summary = BUSCO_QC.out.summary
             ch_versions = ch_versions.mix(BUSCO_QC.out.versions.first())