Add module muse/sump (nf-core#5629)

* start work on muse/sump * update to include tbi * uodate snap * update paths * prettier * exclude conda test * change to bioconda * update meta and snap * add conda test * update tests * change ext args to new convention * update containers and version extraction and add gzip * add bgzip * Update modules/nf-core/muse/sump/meta.yml Co-authored-by: Simon Pearce <[email protected]> * update test * finalize bgzip and specify tests * fix meta --------- Co-authored-by: Simon Pearce <[email protected]>
LouisLeNezet · Dec 4, 2024 · 2ebe102 · 2ebe102
1 parent f6c776a
commit 2ebe102
Show file tree

Hide file tree

Showing 7 changed files with 268 additions and 0 deletions.
diff --git a/modules/nf-core/muse/sump/environment.yml b/modules/nf-core/muse/sump/environment.yml
@@ -0,0 +1,6 @@
+channels:
+  - conda-forge
+  - bioconda
+dependencies:
+  - bioconda::muse=2.1.2
+  - bioconda::tabix=1.11 # needed for bgzip
diff --git a/modules/nf-core/muse/sump/main.nf b/modules/nf-core/muse/sump/main.nf
@@ -0,0 +1,54 @@
+process MUSE_SUMP {
+    tag "$meta.id"
+    label 'process_medium'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://community-cr-prod.seqera.io/docker/registry/v2/blobs/sha256/35/3567f6162ff718c648175c5e7b5f848eaa27811d0cb3ad53def8f0a1c8893efa/data':
+        'community.wave.seqera.io/library/muse_tabix:df58ca78bd9447b7' }"
+
+    input:
+    tuple val(meta), path(muse_call_txt)
+    tuple val(meta2), path(ref_vcf), path(ref_vcf_tbi)
+
+    output:
+    tuple val(meta), path("*.vcf.gz"), emit: vcf
+    path "versions.yml"              , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args   = task.ext.args   ?: '' // -G for WGS data and -E for WES data
+    def args2  = task.ext.args2  ?: '' // args for bgzip
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    MuSE \\
+        sump \\
+        $args \\
+        -I $muse_call_txt \\
+        -n $task.cpus    \\
+        -D $ref_vcf \\
+        -O ${prefix}.vcf
+
+    bgzip $args2 --threads $task.cpus ${prefix}.vcf
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        MuSE: \$( MuSE --version | sed -e "s/MuSE, version //g" )
+        bgzip: \$( bgzip --version | sed -e "s/bgzip (htslib) //g" )
+    END_VERSIONS
+    """
+
+    stub:
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    echo "" | gzip > ${prefix}.vcf.gz
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        MuSE: \$( MuSE --version | sed -e "s/MuSE, version //g" )
+        bgzip: \$( bgzip --version | sed -e "s/bgzip (htslib) //g" )
+    END_VERSIONS
+    """
+}
diff --git a/modules/nf-core/muse/sump/meta.yml b/modules/nf-core/muse/sump/meta.yml
@@ -0,0 +1,67 @@
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/meta-schema.json
+name: "muse_sump"
+description: Computes tier-based cutoffs from a sample-specific error model which
+  is generated by muse/call and reports the finalized variants
+keywords:
+  - variant calling
+  - somatic
+  - wgs
+  - wxs
+  - vcf
+tools:
+  - "MuSE":
+      description: "Somatic point mutation caller based on Markov substitution model
+        for molecular evolution"
+      homepage: "https://bioinformatics.mdanderson.org/public-software/muse/"
+      documentation: "https://github.com/wwylab/MuSE"
+      tool_dev_url: "https://github.com/wwylab/MuSE"
+      doi: "10.1101/gr.278456.123"
+      licence: ["https://github.com/danielfan/MuSE/blob/master/LICENSE"]
+      identifier: ""
+
+input:
+  - - meta:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'sample1', single_end:false ]`
+    - muse_call_txt:
+        type: file
+        description: single input file generated by 'MuSE call'
+        pattern: "*.MuSE.txt"
+  - - meta2:
+        type: map
+        description: |
+          Groovy Map containing reference information.
+          e.g. `[ id:'test' ]`
+    - ref_vcf:
+        type: file
+        description: |
+          dbSNP vcf file that should be bgzip compressed, tabix indexed and
+          based on the same reference genome used in 'MuSE call'
+        pattern: ".vcf.gz"
+    - ref_vcf_tbi:
+        type: file
+        description: Tabix index for the dbSNP vcf file
+        pattern: ".vcf.gz.tbi"
+output:
+  - vcf:
+      - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. `[ id:'sample1' ]`
+          pattern: "*.vcf"
+      - "*.vcf.gz":
+          type: map
+          description: bgzipped vcf file with called variants
+          pattern: "*.vcf"
+  - versions:
+      - versions.yml:
+          type: file
+          description: File containing software versions
+          pattern: "versions.yml"
+authors:
+  - "@famosab"
+maintainers:
+  - "@famosab"
diff --git a/modules/nf-core/muse/sump/tests/main.nf.test b/modules/nf-core/muse/sump/tests/main.nf.test
@@ -0,0 +1,78 @@
+nextflow_process {
+
+    name "Test Process MUSE_SUMP"
+    script "../main.nf"
+    process "MUSE_SUMP"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "muse"
+    tag "muse/sump"
+
+    test("human - txt") {
+
+        config "./nextflow.config"
+
+        when {
+            params {
+                module_args = '-E'
+            }
+            process {
+                """
+                input[0] = [
+                    [ id:'test' ], // meta map
+                    file(params.modules_testdata_base_path + 'genomics/homo_sapiens/illumina/muse/MuSE-call.chr21.hg38.paired_end.recal.MuSE.txt', checkIfExists: true)
+                ]
+                input[1] = [
+                    [ id:'reference' ], // meta map
+                    file(params.modules_testdata_base_path + 'genomics/homo_sapiens/genome/chr21/germlineresources/dbsnp_138.hg38.vcf.gz', checkIfExists: true),
+                    file(params.modules_testdata_base_path + 'genomics/homo_sapiens/genome/chr21/germlineresources/dbsnp_138.hg38.vcf.gz.tbi', checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(
+                    process.out.versions,
+                    path(process.out.vcf.get(0).get(1)).vcf.header.getColumnCount(),
+                    path(process.out.vcf.get(0).get(1)).vcf.summary
+                    ).match()
+                }
+            )
+        }
+
+    }
+
+    test("human - txt - stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id:'test' ], // meta map
+                    file(params.modules_testdata_base_path + 'genomics/homo_sapiens/illumina/muse/MuSE-call.chr21.hg38.paired_end.recal.MuSE.txt', checkIfExists: true)
+                ]
+                input[1] = [
+                    [ id:'reference' ], // meta map
+                    file(params.modules_testdata_base_path + 'genomics/homo_sapiens/genome/chr21/germlineresources/dbsnp_138.hg38.vcf.gz', checkIfExists: true),
+                    file(params.modules_testdata_base_path + 'genomics/homo_sapiens/genome/chr21/germlineresources/dbsnp_138.hg38.vcf.gz.tbi', checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+}
diff --git a/modules/nf-core/muse/sump/tests/main.nf.test.snap b/modules/nf-core/muse/sump/tests/main.nf.test.snap
@@ -0,0 +1,49 @@
+{
+    "human - txt - stub": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.vcf.gz:md5,68b329da9893e34099c7d8ad5cb9c940"
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,4d667cb8f2f96c5705b1e44affdd7330"
+                ],
+                "vcf": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.vcf.gz:md5,68b329da9893e34099c7d8ad5cb9c940"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,4d667cb8f2f96c5705b1e44affdd7330"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.0"
+        },
+        "timestamp": "2024-11-29T14:52:29.011666"
+    },
+    "human - txt": {
+        "content": [
+            [
+                "versions.yml:md5,4d667cb8f2f96c5705b1e44affdd7330"
+            ],
+            11,
+            "VcfFile [chromosomes=[], sampleCount=2, variantCount=0, phased=true, phasedAutodetect=true]"
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.2"
+        },
+        "timestamp": "2024-12-02T16:29:22.506313"
+    }
+}
diff --git a/modules/nf-core/muse/sump/tests/nextflow.config b/modules/nf-core/muse/sump/tests/nextflow.config
@@ -0,0 +1,12 @@
+process {
+
+    withName: 'MUSE_SUMP' {
+        ext.args    = params.module_args
+        stageInMode = 'copy'
+    }
+
+    withName: 'MUSE_SUMP_WGS' {
+        ext.args    = '-G'
+    }
+
+}
diff --git a/modules/nf-core/muse/sump/tests/tags.yml b/modules/nf-core/muse/sump/tests/tags.yml
@@ -0,0 +1,2 @@
+muse/sump:
+  - "modules/nf-core/muse/sump/**"