Merge pull request #1714 from nvnieuwk/new-module-gatk4/ComposeSTRTableFile

added gatk4/composestrtablefile

modules/gatk4/composestrtablefile/main.nf

@@ -0,0 +1,53 @@
+process GATK4_COMPOSESTRTABLEFILE {
+    tag "$fasta"
+    label 'process_low'
+
+    conda (params.enable_conda ? "bioconda::gatk4=4.2.6.1" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/gatk4:4.2.6.1--hdfd78af_0':
+        'quay.io/biocontainers/gatk4:4.2.6.1--hdfd78af_0' }"
+
+    input:
+    path(fasta)
+    path(fasta_fai)
+    path(dict)
+
+    output:
+    path "*.zip"            , emit: str_table
+    path "versions.yml"     , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+
+    def avail_mem = 6
+    if (!task.memory) {
+        log.info '[GATK ComposeSTRTableFile] Available memory not known - defaulting to 6GB. Specify process memory requirements to change this.'
+    } else {
+        avail_mem = task.memory.giga
+    }
+    """
+    gatk --java-options "-Xmx${avail_mem}g" ComposeSTRTableFile \\
+        --reference $fasta \\
+        --output ${fasta.baseName}.zip \\
+        --tmp-dir . \\
+        $args
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        gatk4: \$(echo \$(gatk --version 2>&1) | sed 's/^.*(GATK) v//; s/ .*\$//')
+    END_VERSIONS
+    """
+
+    stub:
+    """
+    touch test.zip
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        gatk4: \$(echo \$(gatk --version 2>&1) | sed 's/^.*(GATK) v//; s/ .*\$//')
+    END_VERSIONS
+    """
+}

modules/gatk4/composestrtablefile/meta.yml

@@ -0,0 +1,43 @@
+name: "gatk4_composestrtablefile"
+description: This tool looks for low-complexity STR sequences along the reference that are later used to estimate the Dragstr model during single sample auto calibration CalibrateDragstrModel.
+keywords:
+  - gatk4
+  - composestrtablefile
+tools:
+  - gatk4:
+      description:
+        Genome Analysis Toolkit (GATK4). Developed in the Data Sciences Platform at the Broad Institute, the toolkit offers a wide variety of tools
+        with a primary focus on variant discovery and genotyping. Its powerful processing engine
+        and high-performance computing features make it capable of taking on projects of any size.
+      homepage: https://gatk.broadinstitute.org/hc/en-us
+      documentation: https://gatk.broadinstitute.org/hc/en-us/articles/4405451249819-ComposeSTRTableFile
+      tool_dev_url: https://github.com/broadinstitute/gatk
+      doi: 10.1158/1538-7445.AM2017-3590
+      licence: ["Apache-2.0"]
+
+input:
+  - fasta:
+      type: file
+      description: FASTA reference file
+      pattern: "*.{fasta,fa}"
+  - fasta_fai:
+      type: file
+      description: index of the FASTA reference file
+      pattern: "*.fai"
+  - dict:
+      type: file
+      description: Sequence dictionary of the FASTA reference file
+      pattern: "*.dict"
+
+output:
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+  - str_table:
+      type: file
+      description: A zipped folder containing the STR table files
+      pattern: "*.zip"
+
+authors:
+  - "@nvnieuwk"

tests/config/pytest_modules.yml

@@ -751,6 +751,10 @@ gatk4/combinegvcfs:
   - modules/gatk4/combinegvcfs/**
   - tests/modules/gatk4/combinegvcfs/**
 
+gatk4/composestrtablefile:
+  - modules/gatk4/composestrtablefile/**
+  - tests/modules/gatk4/composestrtablefile/**
+
 gatk4/createsequencedictionary:
   - modules/gatk4/createsequencedictionary/**
   - tests/modules/gatk4/createsequencedictionary/**

tests/modules/gatk4/composestrtablefile/main.nf

@@ -0,0 +1,16 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { GATK4_COMPOSESTRTABLEFILE } from '../../../../modules/gatk4/composestrtablefile/main.nf'
+
+workflow test_gatk4_composestrtablefile {
+    
+    fasta = file(params.test_data['homo_sapiens']['genome']['genome_fasta'], checkIfExists: true)
+
+    fasta_fai = file(params.test_data['homo_sapiens']['genome']['genome_fasta_fai'], checkIfExists: true)
+
+    dict = file(params.test_data['homo_sapiens']['genome']['genome_dict'], checkIfExists: true)
+
+    GATK4_COMPOSESTRTABLEFILE ( fasta, fasta_fai, dict )
+}

tests/modules/gatk4/composestrtablefile/nextflow.config

@@ -0,0 +1,5 @@
+process {
+
+    publishDir = { "${params.outdir}/${task.process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()}" }
+    
+}

tests/modules/gatk4/composestrtablefile/test.yml

@@ -0,0 +1,7 @@
+- name: gatk4 composestrtablefile test_gatk4_composestrtablefile
+  command: nextflow run ./tests/modules/gatk4/composestrtablefile -entry test_gatk4_composestrtablefile -c ./tests/config/nextflow.config  -c ./tests/modules/gatk4/composestrtablefile/nextflow.config
+  tags:
+    - gatk4/composestrtablefile
+    - gatk4
+  files:
+    - path: output/gatk4/genome.zip