Merge branch 'master' into 2FAST2Q

nf-core · Jan 24, 2025 · b350d66 · b350d66
2 parents f114f57 + 271b0be
commit b350d66
Show file tree

Hide file tree

Showing 10 changed files with 429 additions and 4 deletions.
diff --git a/modules/nf-core/catpack/addnames/meta.yml b/modules/nf-core/catpack/addnames/meta.yml
@@ -53,13 +53,11 @@ output:
           description: |
             Groovy Map containing sample information
             e.g. `[ id:'sample1', single_end:false ]`
-          pattern: "*.{bam,cram,sam}"
 
       - ${prefix}.txt:
           type: map
           description: |
-            Groovy Map containing sample information
-            e.g. `[ id:'sample1', single_end:false ]`
+            CAT/BAT/RAT classification file with added taxonomic names
           pattern: "*.{bam,cram,sam}"
           ontologies:
             - edam: "http://edamontology.org/format_3475"

diff --git a/modules/nf-core/catpack/bins/tests/nextflow.config b/modules/nf-core/catpack/bins/tests/nextflow.config
@@ -6,4 +6,8 @@ process {
     withName: CATPACK_BINS {
         ext.args = "--bin_suffix .fasta"
     }
+
+    withName: CATPACK_ADDNAMES {
+        ext.args = "--only_official"
+    }
 }
diff --git a/modules/nf-core/catpack/download/main.nf b/modules/nf-core/catpack/download/main.nf
@@ -37,6 +37,12 @@ process CATPACK_DOWNLOAD {
     def args = task.ext.args ?: ''
     prefix = task.ext.prefix ?: "${meta.id}"
     """
+    echo "CAT_pack \\
+        download \\
+        ${args} \\
+        --db ${db}
+        -o ${prefix}/"
+
     mkdir ${prefix}/
 
     cat <<-END_VERSIONS > versions.yml

diff --git a/modules/nf-core/catpack/summarise/environment.yml b/modules/nf-core/catpack/summarise/environment.yml
@@ -0,0 +1,7 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/environment-schema.json
+channels:
+  - conda-forge
+  - bioconda
+dependencies:
+  - "bioconda::cat=6.0.1"
diff --git a/modules/nf-core/catpack/summarise/main.nf b/modules/nf-core/catpack/summarise/main.nf
@@ -0,0 +1,62 @@
+process CATPACK_SUMMARISE {
+    tag "${meta.id}"
+    label 'process_single'
+
+    conda "${moduleDir}/environment.yml"
+    container "${workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container
+        ? 'https://depot.galaxyproject.org/singularity/cat:6.0.1--hdfd78af_0'
+        : 'biocontainers/cat:6.0.1--hdfd78af_0'}"
+
+    input:
+    tuple val(meta), path(classification)
+    tuple val(meta2), path(contigs)
+
+    output:
+    tuple val(meta), path("*.txt"), emit: txt
+    path "versions.yml", emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    if ("${classification}" == "${prefix}.txt") {
+        error("Input and output names are the same, set prefix in module configuration to disambiguate!")
+    }
+    def insert_contigs = contigs ? "-c ${contigs}" : ''
+    """
+    CAT_pack summarise \\
+        ${args} \\
+        -i ${classification} \\
+        ${insert_contigs} \\
+        -o ${prefix}.txt
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        catpack: \$(CAT_pack --version | sed 's/CAT_pack pack v//g;s/ .*//g')
+    END_VERSIONS
+    """
+
+    stub:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    if ("${classification}" == "${prefix}.txt") {
+        error("Input and output names are the same, set prefix in module configuration to disambiguate!")
+    }
+    def insert_contigs = contigs ? "-c ${contigs}" : ''
+    """
+    echo "CAT_pack summarise \\
+        ${args} \\
+        -i ${classification} \\
+        ${insert_contigs} \\
+        -o ${prefix}.txt"
+
+    touch ${prefix}.txt
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        catpack: \$(CAT_pack --version | sed 's/CAT_pack pack v//g;s/ .*//g')
+    END_VERSIONS
+    """
+}
diff --git a/modules/nf-core/catpack/summarise/meta.yml b/modules/nf-core/catpack/summarise/meta.yml
@@ -0,0 +1,72 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/meta-schema.json
+name: "catpack_summarise"
+description: Summarises results from CAT/BAT/RAT classification steps
+keywords:
+  - taxonomic classification
+  - classification
+  - long reads
+  - mags
+  - assembly
+tools:
+  - "catpack":
+      description: "CAT/BAT: tool for taxonomic classification of contigs and metagenome-assembled
+        genomes (MAGs)"
+      homepage: "https://github.com/MGXlab/CAT_pack"
+      documentation: "https://github.com/MGXlab/CAT_pack"
+      tool_dev_url: "https://github.com/MGXlab/CAT_pack"
+      doi: "10.1186/s13059-019-1817-x"
+      licence: ["MIT"]
+      identifier: ""
+
+input:
+  - - meta:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'sample1', single_end:false ]`
+
+    - classification:
+        type: file
+        description: CAT/BAT/RAT classification table annotated with official names (from CAT_pack addnames)
+        pattern: "*.txt"
+        ontologies:
+          - edam: "http://edamontology.org/format_3475"
+  - - meta2:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'sample1', single_end:false ]`
+
+    - contigs:
+        type: file
+        description: Optional nucleotide FASTA file containing long DNA sequences such as contigs that were classified (only if classification table is from CAT_pack contigs)
+        pattern: "*.{fasta,fna,fa,fas}"
+        ontologies:
+          - edam: "http://edamontology.org/format_1929"
+
+output:
+  - txt:
+      - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. `[ id:'sample1', single_end:false ]`
+      - "*.txt":
+          type: file
+          description: |
+            Summary statistics table of CAT/BAT/RAT results
+          pattern: "*.txt"
+          ontologies:
+            - edam: "http://edamontology.org/format_3475"
+
+  - versions:
+      - "versions.yml":
+          type: file
+          description: File containing software versions
+          pattern: "versions.yml"
+
+authors:
+  - "@jfy133"
+maintainers:
+  - "@jfy133"
diff --git a/modules/nf-core/catpack/summarise/tests/main.nf.test b/modules/nf-core/catpack/summarise/tests/main.nf.test
@@ -0,0 +1,158 @@
+nextflow_process {
+
+    name "Test Process CATPACK_SUMMARISE"
+    script "../main.nf"
+    process "CATPACK_SUMMARISE"
+    config './nextflow.config'
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "catpack"
+    tag "catpack/summarise"
+    tag "catpack/prepare"
+    tag "catpack/contigs"
+    tag "catpack/bins"
+    tag "catpack/addnames"
+
+    setup {
+        run('CATPACK_PREPARE') {
+            script '../../prepare/main.nf'
+            process {
+            """
+            input[0] = [ [id:'test'], [ file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/proteome.fasta', checkIfExists: true)  ] ]
+            input[1] = [ file(params.modules_testdata_base_path + 'genomics/sarscov2/metagenome/prot_nodes.dmp', checkIfExists: true) ]
+            input[2] = [ file(params.modules_testdata_base_path + 'genomics/sarscov2/metagenome/prot_names.dmp', checkIfExists: true) ]
+            input[3] = [ file(params.modules_testdata_base_path + 'genomics/sarscov2/metagenome/prot.accession2taxid.gz', checkIfExists: true) ]
+            """
+            }
+        }
+    }
+
+    test("sarscov2 - contigs - with names") {
+
+        setup {
+            run('CATPACK_CONTIGS') {
+                script '../../contigs/main.nf'
+                process {
+                """
+                    input[0] = [ [id:'test'], [ file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)  ] ]
+                    input[1] = CATPACK_PREPARE.out.db
+                    input[2] = CATPACK_PREPARE.out.taxonomy
+                    input[3] = [[:], []]
+                    input[4] = [[:], []]
+                """
+                }
+            }
+
+            run('CATPACK_ADDNAMES') {
+                    script '../../addnames/main.nf'
+                    process {
+                    """
+                    input[0] = CATPACK_CONTIGS.out.contig2classification
+                    input[1] = CATPACK_PREPARE.out.taxonomy
+                    """
+                    }
+                }
+        }
+
+        when {
+            process {
+                """
+                input[0] = CATPACK_ADDNAMES.out.txt
+                input[1] = [ [id:'test'], [ file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)  ] ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+    test("sarscov2 - bins - raw") {
+
+        setup {
+            run('CATPACK_BINS') {
+                    script '../../bins/main.nf'
+                    process {
+                    """
+                    input[0] = [ [id:'test'], [ file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)  ] ]
+                    input[1] = CATPACK_PREPARE.out.db
+                    input[2] = CATPACK_PREPARE.out.taxonomy
+                    input[3] = [[:], []]
+                    input[4] = [[:], []]
+                    """
+                    }
+                }
+
+            run('CATPACK_ADDNAMES') {
+                    script '../../addnames/main.nf'
+                    process {
+                    """
+                    input[0] = CATPACK_BINS.out.bin2classification
+                    input[1] = CATPACK_PREPARE.out.taxonomy
+                    """
+                    }
+                }
+        }
+
+        when {
+            process {
+                """
+                input[0] = CATPACK_ADDNAMES.out.txt
+                input[1] = [[:],[]]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+    test("sarscov2 - contigs - stub") {
+
+        options "-stub"
+
+        setup {
+            run('CATPACK_CONTIGS') {
+                script '../../contigs/main.nf'
+                process {
+                """
+                    input[0] = [ [id:'test'], [ file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)  ] ]
+                    input[1] = CATPACK_PREPARE.out.db
+                    input[2] = CATPACK_PREPARE.out.taxonomy
+                    input[3] = [[:], []]
+                    input[4] = [[:], []]
+                """
+                }
+            }
+        }
+
+        when {
+            process {
+                """
+                input[0] = CATPACK_CONTIGS.out.contig2classification
+                input[1] = [ [id:'test'], [ file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)  ] ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+}