Merge pull request #366 from prototaxites/cat_summarise

Add CAT_SUMMARY process and offical_taxonomy param
nf-core · Dec 16, 2022 · 9ad1729 · 9ad1729
2 parents c3a0ef3 + 7b0adf6
commit 9ad1729
Show file tree

Hide file tree

Showing 8 changed files with 55 additions and 3 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -10,6 +10,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - [#353](https://github.com/nf-core/mag/pull/353) - Added the busco_clean parameter to optionally clean each BUSCO directory after a successful
 - [#361](https://github.com/nf-core/mag/pull/361) - Added the skip_clipping parameter to skip read preprocessing with fastp or adapterremoval. Running the pipeline with skip_clipping, keep_phix and without specifying a host genome or fasta file skips the FASTQC_TRIMMED process.
 - [#365](https://github.com/nf-core/mag/pull/365) - Adds CONCOCT as an additional (optional) binning tool
+- [#366](https://github.com/nf-core/mag/pull/366) - Added CAT_SUMMARISE process and cat_official_taxonomy parameter.
 
 ### `Changed`
 

diff --git a/conf/modules.config b/conf/modules.config
@@ -284,6 +284,14 @@ process {
             saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
         ]
     }
+    withName: CAT_SUMMARY {
+        ext.prefix = "cat_summary"
+        publishDir = [
+            path: { "${params.outdir}/Taxonomy/CAT/" },
+            mode: params.publish_dir_mode,
+            saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+        ]
+    }
 
     withName: GTDBTK_CLASSIFY {
         ext.args   = "--extension fa"

diff --git a/conf/test_full.config b/conf/test_full.config
@@ -30,4 +30,7 @@ params {
     megahit_fix_cpu_1     = true
     // available options to enable reproducibility for BUSCO (--busco_download_path or --busco_reference) not used here
     // to allow detection of possible problems in automated lineage selection mode using public databases
+
+    // test CAT with official taxonomic ranks only
+    cat_official_taxonomy         = true
 }
diff --git a/modules/local/cat.nf b/modules/local/cat.nf
@@ -20,10 +20,11 @@ process CAT {
     path "versions.yml"                    , emit: versions
 
     script:
+    def official_taxonomy = params.cat_official_taxonomy ? "--official_taxonomy" : ""
     """
     CAT bins -b "bins/" -d database/ -t taxonomy/ -n "${task.cpus}" -s .fa --top 6 -o "${meta.assembler}-${meta.binner}-${meta.id}" --I_know_what_Im_doing
-    CAT add_names -i "${meta.assembler}-${meta.binner}-${meta.id}.ORF2LCA.txt" -o "${meta.assembler}-${meta.binner}-${meta.id}.ORF2LCA.names.txt" -t taxonomy/
-    CAT add_names -i "${meta.assembler}-${meta.binner}-${meta.id}.bin2classification.txt" -o "${meta.assembler}-${meta.binner}-${meta.id}.bin2classification.names.txt" -t taxonomy/
+    CAT add_names -i "${meta.assembler}-${meta.binner}-${meta.id}.ORF2LCA.txt" -o "${meta.assembler}-${meta.binner}-${meta.id}.ORF2LCA.names.txt" -t taxonomy/ ${official_taxonomy}
+    CAT add_names -i "${meta.assembler}-${meta.binner}-${meta.id}.bin2classification.txt" -o "${meta.assembler}-${meta.binner}-${meta.id}.bin2classification.names.txt" -t taxonomy/ ${official_taxonomy}
 
     mkdir raw
     mv *.ORF2LCA.txt *.predicted_proteins.faa *.predicted_proteins.gff *.log *.bin2classification.txt raw/

diff --git a/modules/local/cat_summary.nf b/modules/local/cat_summary.nf
@@ -0,0 +1,29 @@
+process CAT_SUMMARY {
+    label 'process_low'
+
+    conda (params.enable_conda ? "bioconda::bioawk=1.0" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/bioawk:1.0--hed695b0_5' :
+        'quay.io/biocontainers/bioawk:1.0--hed695b0_5' }"
+
+    input:
+    path(cat_summaries)
+
+    output:
+    path("*.tsv")      , emit: combined
+    path "versions.yml", emit: versions
+
+    script:
+    def prefix = task.ext.prefix ?: "cat_summary"
+    """
+    # use find as sometimes these are empty and need to fail gracefully
+    find -L -type f -name "*bin2classification.names.txt.gz" -exec sh -c 'for f do gunzip -c \$f > \${f%.*}; done' find-sh {} +
+
+    bioawk '(NR == 1) || (FNR > 1)' *.txt > ${prefix}.tsv
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        bioawk: \$(bioawk --version | cut -f 3 -d ' ' )
+    END_VERSIONS
+    """
+}
diff --git a/nextflow.config b/nextflow.config
@@ -67,6 +67,7 @@ params {
     skip_krona                           = false
     cat_db                               = null
     cat_db_generate                      = false
+    cat_official_taxonomy                = false
     save_cat_db                          = false
     gtdb                                 = "https://data.ace.uq.edu.au/public/gtdb/data/releases/release202/202.0/auxillary_files/gtdbtk_r202_data.tar.gz"
     gtdbtk_min_completeness              = 50.0

diff --git a/nextflow_schema.json b/nextflow_schema.json
@@ -447,6 +447,10 @@
                     "description": "Save the CAT database generated when specified by `--cat_db_generate`.",
                     "help_text": "Useful to allow reproducibility, as old versions of prebuild CAT databases do not always remain accessible and underlying NCBI taxonomy and nr databases change."
                 },
+                "cat_official_taxonomy": {
+                    "type": "boolean",
+                    "description": "Only return official taxonomic ranks (Kingdom, Phylum, etc.) when running CAT."
+                },
                 "gtdb": {
                     "type": "string",
                     "default": "https://data.gtdb.ecogenomic.org/releases/release202/202.0/auxillary_files/gtdbtk_r202_data.tar.gz",
@@ -641,7 +645,7 @@
                     "type": "number",
                     "default": 0.5,
                     "description": "Specify single-copy gene score threshold for bin refinement.",
-                    "help_text": "Score threshold for single-copy gene selection algorithm to keep selecting bins, with a value ranging from 0-1.\n\nFor description of scoring algorithm, see: Sieber, Christian M. K., et al. 2018. Nature Microbiology 3 (7): 836\u201343. https://doi.org/10.1038/s41564-018-0171-1.\n\n> Modifies DAS Tool parameter --score_threshold\n"
+                    "help_text": "Score threshold for single-copy gene selection algorithm to keep selecting bins, with a value ranging from 0-1.\n\nFor description of scoring algorithm, see: Sieber, Christian M. K., et al. 2018. Nature Microbiology 3 (7): 836–43. https://doi.org/10.1038/s41564-018-0171-1.\n\n> Modifies DAS Tool parameter --score_threshold\n"
                 },
                 "postbinning_input": {
                     "type": "string",

diff --git a/workflows/mag.nf b/workflows/mag.nf
@@ -83,6 +83,7 @@ include { QUAST_BINS_SUMMARY                                  } from '../modules
 include { CAT_DB                                              } from '../modules/local/cat_db'
 include { CAT_DB_GENERATE                                     } from '../modules/local/cat_db_generate'
 include { CAT                                                 } from '../modules/local/cat'
+include { CAT_SUMMARY                                         } from "../modules/local/cat_summary"
 include { BIN_SUMMARY                                         } from '../modules/local/bin_summary'
 include { COMBINE_TSV                                         } from '../modules/local/combine_tsv'
 include { MULTIQC                                             } from '../modules/local/multiqc'
@@ -640,7 +641,11 @@ workflow MAG {
             ch_input_for_postbinning_bins,
             ch_cat_db
         )
+        CAT_SUMMARY(
+            CAT.out.tax_classification.collect()
+        )
         ch_versions = ch_versions.mix(CAT.out.versions.first())
+        ch_versions = ch_versions.mix(CAT_SUMMARY.out.versions.first())
 
         /*
          * GTDB-tk: taxonomic classifications using GTDB reference