Merge pull request #1655 from nvnieuwk/new-module-gatk4/splitintervals

new module: gatk4/splitintervals

modules/gatk4/splitintervals/main.nf

@@ -0,0 +1,48 @@
+process GATK4_SPLITINTERVALS {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda (params.enable_conda ? "bioconda::gatk4=4.2.6.1" : null)
+    container "${ 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:
+    tuple val(meta), path(intervals)
+    path(fasta)
+    path(fasta_fai)
+    path(dict)
+
+    output:
+    tuple val(meta), path("**.interval_list"), emit: split_intervals
+    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}"
+    def reference = fasta ? "--reference $fasta" : ""
+
+    def avail_mem = 3
+    if (!task.memory) {
+        log.info '[GATK SplitIntervals] Available memory not known - defaulting to 3GB. Specify process memory requirements to change this.'
+    } else {
+        avail_mem = task.memory.giga
+    }
+
+    """
+    gatk --java-options "-Xmx${avail_mem}g" SplitIntervals \\
+        --output ${prefix} \\
+        --intervals $intervals \\
+        $reference \\
+        --tmp-dir . \\
+        $args
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        gatk4: \$(echo \$(gatk --version 2>&1) | sed 's/^.*(GATK) v//; s/ .*\$//')
+    END_VERSIONS
+    """
+}

modules/gatk4/splitintervals/meta.yml

@@ -0,0 +1,53 @@
+name: gatk4_splitintervals
+keywords:
+  - interval
+  - bed
+tools:
+  - gatk4:
+      description: Genome Analysis Toolkit (GATK4)
+      homepage: https://gatk.broadinstitute.org/hc/en-us
+      documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
+      tool_dev_url: https://github.com/broadinstitute/gatk
+      doi: "10.1158/1538-7445.AM2017-3590"
+      licence: ["BSD-3-clause"]
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test' ]
+  - interval:
+      type: file
+      description: Interval list or BED
+      pattern: "*.{interval,interval_list,bed}"
+  - fasta:
+      type: file
+      description: Reference FASTA
+      pattern: "*.{fa,fasta}"
+  - fasta_fai:
+      type: file
+      description: Reference FASTA index
+      pattern: "*.fai"
+  - dict:
+      type: file
+      description: Reference sequence dictionary
+      pattern: "*.dict"
+
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test' ]
+  - bed:
+      type: file
+      description: A list of scattered interval lists
+      pattern: "*.interval_list"
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+
+authors:
+  - "@nvnieuwk"

tests/config/pytest_modules.yml

@@ -819,6 +819,10 @@ gatk4/selectvariants:
   - modules/gatk4/selectvariants/**
   - tests/modules/gatk4/selectvariants/**
 
+gatk4/splitintervals:
+  - modules/gatk4/splitintervals/**
+  - tests/modules/gatk4/splitintervals/**
+
 gatk4/splitncigarreads:
   - modules/gatk4/splitncigarreads/**
   - tests/modules/gatk4/splitncigarreads/**

tests/modules/gatk4/splitintervals/main.nf

@@ -0,0 +1,33 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { GATK4_SPLITINTERVALS } from '../../../../modules/gatk4/splitintervals/main.nf'
+
+workflow test_gatk4_splitintervals_bed {
+    
+    input = [
+        [ id:'test' ], // meta map
+        file(params.test_data['homo_sapiens']['genome']['genome_multi_interval_bed'], checkIfExists: true)
+    ]
+
+    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)
+    fasta_dict = file(params.test_data['homo_sapiens']['genome']['genome_dict'], checkIfExists: true)
+
+    GATK4_SPLITINTERVALS ( input, fasta, fasta_fai, fasta_dict)
+}
+
+workflow test_gatk4_splitintervals_intervals {
+    
+    input = [
+        [ id:'test' ], // meta map
+        file(params.test_data['homo_sapiens']['genome']['genome_interval_list'], checkIfExists: true)
+    ]
+
+    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)
+    fasta_dict = file(params.test_data['homo_sapiens']['genome']['genome_dict'], checkIfExists: true)
+
+    GATK4_SPLITINTERVALS ( input, fasta, fasta_fai, fasta_dict)
+}

tests/modules/gatk4/splitintervals/nextflow.config

@@ -0,0 +1,9 @@
+process {
+
+    publishDir = { "${params.outdir}/${task.process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()}" }
+    
+
+    withName: GATK4_SPLITINTERVALS {
+        ext.args = "--scatter-count 2"
+    }
+}

tests/modules/gatk4/splitintervals/test.yml

@@ -0,0 +1,23 @@
+- name: gatk4 splitintervals test_gatk4_splitintervals_bed
+  command: nextflow run tests/modules/gatk4/splitintervals -entry test_gatk4_splitintervals_bed -c tests/config/nextflow.config
+  tags:
+    - gatk4/splitintervals
+    - gatk4
+  files:
+    - path: output/gatk4/test/0000-scattered.interval_list
+      md5sum: c8d6b19e7a92535b6ce9608eae558faa
+    - path: output/gatk4/test/0001-scattered.interval_list
+      md5sum: b1877ad96aec308906594c50ebbe3ded
+    - path: output/gatk4/versions.yml
+
+- name: gatk4 splitintervals test_gatk4_splitintervals_intervals
+  command: nextflow run tests/modules/gatk4/splitintervals -entry test_gatk4_splitintervals_intervals -c tests/config/nextflow.config
+  tags:
+    - gatk4/splitintervals
+    - gatk4
+  files:
+    - path: output/gatk4/test/0000-scattered.interval_list
+      md5sum: ebd6b34a335efc6732ff541936c6d2d5
+    - path: output/gatk4/test/0001-scattered.interval_list
+      md5sum: 9459b0e124fa84ec1e64ac4615bc9af7
+    - path: output/gatk4/versions.yml