new module : jvarkit/vcffilterjdk (nf-core#6621)

* vcffilterdjdk * update params * update params * oppsss tag and TODO * target/region * answers to review * f...g space * fix conda problem https://nfcore.slack.com/archives/CJRH30T6V/p1726233311260959 * add test+bed * reset polyx * prevent test exception md5sum for empty file * update main.nf.test * update meta.yml * remove suggestion --------- Co-authored-by: James A. Fellows Yates <[email protected]>
herpov · Oct 2, 2024 · af02735 · af02735
1 parent 6866c4e
commit af02735
Show file tree

Hide file tree

Showing 7 changed files with 385 additions and 0 deletions.
diff --git a/modules/nf-core/jvarkit/vcffilterjdk/environment.yml b/modules/nf-core/jvarkit/vcffilterjdk/environment.yml
@@ -0,0 +1,8 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/environment-schema.json
+channels:
+  - conda-forge
+  - bioconda
+dependencies:
+  - "bioconda::jvarkit=2024.08.25"
+  - "bioconda:bcftools=1.20"
diff --git a/modules/nf-core/jvarkit/vcffilterjdk/main.nf b/modules/nf-core/jvarkit/vcffilterjdk/main.nf
@@ -0,0 +1,88 @@
+process JVARKIT_VCFFILTERJDK {
+    tag "$meta.id"
+    label 'process_single'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/jvarkit:2024.08.25--hdfd78af_1':
+        'biocontainers/jvarkit:2024.08.25--hdfd78af_1' }"
+
+    input:
+    tuple val(meta),  path(vcf), path(tbi), path(regions_file)
+    tuple val(meta2), path(fasta)
+    tuple val(meta3), path(fai)
+    tuple val(meta4), path(dict)
+    tuple val(meta5), path(code)
+    tuple val(meta6), path(pedigree)
+
+    output:
+    tuple val(meta), path("*.${extension}"), emit: vcf
+    tuple val(meta), path("*.tbi")         , emit: tbi, optional: true
+    tuple val(meta), path("*.csi")         , emit: csi, optional: true
+    path "versions.yml"                    , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args1         = task.ext.args1 ?: ''
+    def args2         = task.ext.args2 ?: ''
+    def args3         = task.ext.args3 ?: ''
+    def prefix        = task.ext.prefix ?: "${meta.id}"
+    def script_file   = code ? "--script \"${code}\"" : ""
+    def pedigree_file = pedigree ? " --pedigree \"${pedigree}\" " : ""
+    def regions_cmd   = regions_file ? (tbi ? " --regions-file" : " --targets-file") + " \"${regions_file}\" " : ""
+
+    extension  = getVcfExtension(args3); /* custom function, see below */
+
+    if ("$vcf" == "${prefix}.${extension}") error "Input and output names are the same, set prefix in module configuration to disambiguate!"
+    """
+    mkdir -p TMP
+
+    bcftools view \\
+        -O v \\
+        ${regions_cmd} \\
+        ${args1} \\
+        "${vcf}" |\\
+        jvarkit -Xmx${task.memory.giga}g -XX:-UsePerfData -Djava.io.tmpdir=TMP vcffilterjdk \\
+            ${pedigree_file} \\
+            ${script_file} \\
+            ${args2} |\\
+        bcftools view \\
+            --output "${prefix}.${extension}" \\
+            ${args3}
+
+    rm -rf TMP
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        bcftools: \$(bcftools --version 2>&1 | head -n1 | sed 's/^.*bcftools //; s/ .*\$//')
+        jvarkit: \$(jvarkit -v)
+    END_VERSIONS
+    """
+
+    stub:
+    def args3  = task.ext.args3 ?: ''
+    extension  = getVcfExtension(args3); /* custom function, see below */
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    touch "${prefix}.${extension}"
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        bcftools: \$(bcftools --version 2>&1 | head -n1 | sed 's/^.*bcftools //; s/ .*\$//')
+        jvarkit: \$(jvarkit -v)
+    END_VERSIONS
+    """
+}
+
+
+
+// Custom Function to get VCF extension
+String getVcfExtension(String args) {
+    return args.contains("--output-type b") || args.contains("-Ob") ? "bcf.gz" :
+        args.contains("--output-type u") || args.contains("-Ou") ? "bcf" :
+        args.contains("--output-type z") || args.contains("-Oz") ? "vcf.gz" :
+        args.contains("--output-type v") || args.contains("-Ov") ? "vcf" :
+        "vcf";
+}
diff --git a/modules/nf-core/jvarkit/vcffilterjdk/meta.yml b/modules/nf-core/jvarkit/vcffilterjdk/meta.yml
@@ -0,0 +1,118 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/meta-schema.json
+name: "jvarkit_vcffilterjdk"
+description: Filtering VCF with dynamically-compiled java expressions
+keywords:
+  - vcf
+  - bcf
+  - filter
+  - variant
+  - java
+  - script
+tools:
+  - "jvarkit":
+      description: "Java utilities for Bioinformatics."
+      homepage: "https://github.com/lindenb/jvarkit"
+      documentation: "https://jvarkit.readthedocs.io/"
+      tool_dev_url: "https://github.com/lindenb/jvarkit"
+      doi: "10.1093/bioinformatics/btx734 "
+      licence: ["MIT License"]
+      args_id: "$args2"
+
+  - "bcftools":
+      description: |
+        View, subset and filter VCF or BCF files by position and filtering expression. Convert between VCF and BCF
+      homepage: "http://samtools.github.io/bcftools/bcftools.html"
+      documentation: "http://www.htslib.org/doc/bcftools.html"
+      doi: "10.1093/bioinformatics/btp352"
+      licence: ["MIT"]
+      args_id: ["$args1", "$args3"]
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing VCF information
+        e.g. [ id:'test_reference' ]
+  - vcf:
+      type: file
+      description: Input VCF/BCF file
+      pattern: "*.{vcf,bcf,vcf.gz,bcf.gz}"
+  - tbi:
+      type: file
+      description: Optional VCF/BCF index file
+      pattern: "*.{tbi,csi}"
+  - regions_file:
+      type: file
+      description: Optional. Restrict to regions listed in a file
+      pattern: "*.{bed,bed.gz,txt,tsv}"
+  - meta2:
+      type: map
+      description: |
+        Groovy Map containing fasta information
+        e.g. [ id:'test_reference' ]
+  - fasta:
+      type: file
+      description: Fasta reference file
+      pattern: "*.fasta"
+  - meta3:
+      type: map
+      description: |
+        Groovy Map containing fasta.fai information
+        e.g. [ id:'test_reference' ]
+  - fai:
+      type: file
+      description: Fasta file index
+      pattern: "*.fasta.fai"
+  - meta4:
+      type: map
+      description: |
+        Groovy Map containing fasta.dict information
+        e.g. [ id:'test_reference' ]
+  - dict:
+      type: file
+      description: GATK sequence dictionary
+      pattern: "*.dict"
+  - meta5:
+      type: map
+      description: |
+        Groovy Map containing code information
+        e.g. [ id:'test_reference' ]
+  - code:
+      type: file
+      description: File containing custom user code . May be empty if script if provided via `task.ext.args2`.
+      pattern: "*.{code,script,txt,tsv,java,js}"
+  - meta6:
+      type: map
+      description: |
+        Groovy Map containing pedigree information
+        e.g. [ id:'test_reference' ]
+  - pedigree:
+      type: file
+      description: Optional jvarkit pedigree.
+      pattern: "*.{tsv,ped,pedigree}"
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing VCF information
+        e.g. [ id:'test', single_end:false ]
+  - vcf:
+      type: file
+      description: VCF filtered output file
+      pattern: "*.{vcf,bcf,vcf.gz,bcf.gz}"
+  - csi:
+      type: file
+      description: Default VCF file index
+      pattern: "*.csi"
+  - tbi:
+      type: file
+      description: Alternative VCF file index
+      pattern: "*.tbi"
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+authors:
+  - "@lindenb"
+maintainers:
+  - "@lindenb"
diff --git a/modules/nf-core/jvarkit/vcffilterjdk/tests/main.nf.test b/modules/nf-core/jvarkit/vcffilterjdk/tests/main.nf.test
@@ -0,0 +1,119 @@
+// nf-core modules test jvarkit/vcffilterjdk
+nextflow_process {
+
+    name "Test Process JVARKIT_VCFFILTERJDK"
+    script "../main.nf"
+    process "JVARKIT_VCFFILTERJDK"
+    config "./nextflow.config"
+
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "jvarkit"
+    tag "jvarkit/vcffilterjdk"
+
+    test("sarscov2 - vcf") {
+
+        when {
+            process {
+                """
+                input[0] =[
+                    [id:"vcf_test"],
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/vcf/test.vcf', checkIfExists: true),
+                    [],
+                    []
+                ]
+                input[1] = [ [:] , file(params.test_data['sarscov2']['genome']['genome_fasta'], checkIfExists: true) ]
+                input[2] = [ [:] , file(params.test_data['sarscov2']['genome']['genome_fasta_fai'], checkIfExists: true) ]
+                input[3] = [ [:] , file(params.test_data['sarscov2']['genome']['genome_dict'], checkIfExists: true) ]
+                input[4] = [ [] , []]
+                input[5] = [ [] , []]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(
+                    path(process.out.vcf[0][1]).vcf.variantsMD5,
+                    process.out.versions
+                    ).match()
+                }
+            )
+        }
+
+    }
+
+
+
+    test("sarscov2 - vcf+bed") {
+
+        when {
+            process {
+                """
+                input[0] =[
+                    [id:"vcf_test"],
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/vcf/test.vcf', checkIfExists: true),
+                    [],
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/bed/test.bed', checkIfExists: true)
+                ]
+                input[1] = [ [:] , file(params.test_data['sarscov2']['genome']['genome_fasta'], checkIfExists: true) ]
+                input[2] = [ [:] , file(params.test_data['sarscov2']['genome']['genome_fasta_fai'], checkIfExists: true) ]
+                input[3] = [ [:] , file(params.test_data['sarscov2']['genome']['genome_dict'], checkIfExists: true) ]
+                input[4] = [ [] , []]
+                input[5] = [ [] , []]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert file(process.out.vcf[0][1]).exists() },
+                { assert snapshot(process.out.versions).match()
+                }
+            )
+        }
+    }
+
+
+
+
+    test("sarscov2 - vcf - stub") {
+
+        options  "-stub"
+
+        when {
+            process {
+                """
+                input[0] =[
+                    [id:"vcf_test"],
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/vcf/test.vcf', checkIfExists: true),
+                    [],
+                    []
+                ]
+                input[1] = [ [:] , file(params.test_data['sarscov2']['genome']['genome_fasta'], checkIfExists: true) ]
+                input[2] = [ [:] , file(params.test_data['sarscov2']['genome']['genome_fasta_fai'], checkIfExists: true) ]
+                input[3] = [ [:] , file(params.test_data['sarscov2']['genome']['genome_dict'], checkIfExists: true) ]
+                input[4] = [ [] , []]
+                input[5] = [ [] , []]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(
+                    path(process.out.vcf[0][1]),
+                    process.out.versions
+                    ).match()
+                }
+            )
+        }
+
+    }
+
+
+}
diff --git a/modules/nf-core/jvarkit/vcffilterjdk/tests/main.nf.test.snap b/modules/nf-core/jvarkit/vcffilterjdk/tests/main.nf.test.snap
@@ -0,0 +1,45 @@
+{
+    "sarscov2 - vcf": {
+        "content": [
+            "335cdc0f8c403378e1e9d75c41c3736f",
+           [
+           "versions.yml:md5,3601751995727e2ee7102d8ef18e5304"
+           ] 
+        ],
+        "meta": {
+            "nf-test": "0.9.0",
+            "nextflow": "24.04.4"
+        },
+        "timestamp": "2024-09-03T14:00:13.118369362"
+    },
+
+
+    "sarscov2 - vcf+bed": {
+        "content": [
+           [
+           "versions.yml:md5,3601751995727e2ee7102d8ef18e5304"
+           ] 
+        ],
+        "meta": {
+            "nf-test": "0.9.0",
+            "nextflow": "24.04.4"
+        },
+        "timestamp": "2024-09-03T14:00:13.118369362"
+    },
+
+
+    "sarscov2 - vcf - stub": {
+        "content": [
+           "vcf_test.vcf:md5,d41d8cd98f00b204e9800998ecf8427e",
+           [
+           "versions.yml:md5,3601751995727e2ee7102d8ef18e5304"
+           ] 
+        ],
+        "meta": {
+            "nf-test": "0.9.0",
+            "nextflow": "24.04.4"
+        },
+        "timestamp": "2024-09-03T14:00:13.118369362"
+    }
+
+}
diff --git a/modules/nf-core/jvarkit/vcffilterjdk/tests/nextflow.config b/modules/nf-core/jvarkit/vcffilterjdk/tests/nextflow.config
@@ -0,0 +1,5 @@
+process {
+    withName: JVARKIT_VCFFILTERJDK {
+        ext.args2=" --expression 'return variant.getStart()%2==1;' "
+    }
+}
diff --git a/modules/nf-core/jvarkit/vcffilterjdk/tests/tags.yml b/modules/nf-core/jvarkit/vcffilterjdk/tests/tags.yml
@@ -0,0 +1,2 @@
+jvarkit/vcffilterjdk:
+  - "modules/nf-core/jvarkit/vcffilterjdk/**"