Merge pull request #1744 from nvnieuwk/new-module-gatk4/reblockgvcf

new module gatk4/reblockgvcf

modules/gatk4/reblockgvcf/main.nf

@@ -0,0 +1,52 @@
+process GATK4_REBLOCKGVCF {
+    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(gvcf), path(tbi), path(intervals)
+    path fasta
+    path fai
+    path dict
+    path dbsnp
+    path dbsnp_tbi
+
+    output:
+    tuple val(meta), path("*.rb.g.vcf.gz"), path("*.tbi")  , emit: vcf
+    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 dbsnp_command = dbsnp ? "--dbsnp $dbsnp" : ""
+    def interval_command = intervals ? "--intervals $intervals" : ""
+
+    def avail_mem = 3
+    if (!task.memory) {
+        log.info '[GATK ReblockGVCF] Available memory not known - defaulting to 3GB. Specify process memory requirements to change this.'
+    } else {
+        avail_mem = task.memory.giga
+    }
+    """
+    gatk --java-options "-Xmx${avail_mem}g" ReblockGVCF \\
+        --variant $gvcf \\
+        --output ${prefix}.rb.g.vcf.gz \\
+        --reference $fasta \\
+        $dbsnp_command \\
+        $interval_command \\
+        --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/reblockgvcf/meta.yml

@@ -0,0 +1,74 @@
+name: "gatk4_reblockgvcf"
+description: Condenses homRef blocks in a single-sample GVCF
+keywords:
+  - gatk4
+  - reblockgvcf
+  - gvcf
+tools:
+  - gatk4:
+    description: |
+      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/categories/360002369672s
+    doi: 10.1158/1538-7445.AM2017-3590
+    licence: ["Apache-2.0"]
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - gvcf:
+      type: file
+      description: GVCF file created using HaplotypeCaller using the '-ERC GVCF' or '-ERC BP_RESOLUTION' mode
+      pattern: "*.{vcf,gvcf}.gz"
+  - tbi:
+      type: file
+      description: Index of the GVCF file
+      pattern: "*.tbi"
+  - intervals:
+      type: file
+      description: Bed file with the genomic regions included in the library (optional)
+  - fasta:
+      type: file
+      description: The reference fasta file
+      pattern: "*.fasta"
+  - fai:
+      type: file
+      description: Index of reference fasta file
+      pattern: "fasta.fai"
+  - dict:
+      type: file
+      description: GATK sequence dictionary
+      pattern: "*.dict"
+  - dbsnp:
+      type: file
+      description: VCF file containing known sites (optional)
+  - dbsnp_tbi:
+      type: file
+      description: VCF index of dbsnp (optional)
+
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+  - gvcf:
+      type: file
+      description: Filtered GVCF
+      pattern: "*rb.g.vcf.gz"
+  - tbi:
+      type: file
+      description: Index of the filtered GVCF
+      pattern: "*rb.g.vcf.gz.tbi"
+
+authors:
+  - "@nvnieuwk"

tests/config/pytest_modules.yml

@@ -859,6 +859,10 @@ gatk4/mutect2:
   - modules/gatk4/mutect2/**
   - tests/modules/gatk4/mutect2/**
 
+gatk4/reblockgvcf:
+  - modules/gatk4/reblockgvcf/**
+  - tests/modules/gatk4/reblockgvcf/**
+
 gatk4/revertsam:
   - modules/gatk4/revertsam/**
   - tests/modules/gatk4/revertsam/**

tests/modules/gatk4/reblockgvcf/main.nf

@@ -0,0 +1,55 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { GATK4_REBLOCKGVCF } from '../../../../modules/gatk4/reblockgvcf/main.nf'
+
+workflow test_gatk4_reblockgvcf {
+    
+    input = [
+        [ id:'test', single_end:false ], // meta map
+        file(params.test_data['sarscov2']['illumina']['test_vcf_gz'], checkIfExists: true),
+        file(params.test_data['sarscov2']['illumina']['test_vcf_gz_tbi'], checkIfExists: true),
+        []
+    ]
+
+    fasta       = file(params.test_data['sarscov2']['genome']['genome_fasta'], checkIfExists: true)
+    fasta_index = file(params.test_data['sarscov2']['genome']['genome_fasta_fai'], checkIfExists: true)
+    dict        = file(params.test_data['sarscov2']['genome']['genome_dict'], checkIfExists: true)
+
+    GATK4_REBLOCKGVCF ( input, fasta, fasta_index, dict, [], [] )
+}
+
+workflow test_gatk4_reblockgvcf_intervals {
+    
+    input = [
+        [ id:'test', single_end:false ], // meta map
+        file(params.test_data['sarscov2']['illumina']['test_vcf_gz'], checkIfExists: true),
+        file(params.test_data['sarscov2']['illumina']['test_vcf_gz_tbi'], checkIfExists: true),
+        file(params.test_data['sarscov2']['genome']['test_bed'], checkIfExists: true)
+    ]
+
+    fasta       = file(params.test_data['sarscov2']['genome']['genome_fasta'], checkIfExists: true)
+    fasta_index = file(params.test_data['sarscov2']['genome']['genome_fasta_fai'], checkIfExists: true)
+    dict        = file(params.test_data['sarscov2']['genome']['genome_dict'], checkIfExists: true)
+
+    GATK4_REBLOCKGVCF ( input, fasta, fasta_index, dict, [], [] )
+}
+
+workflow test_gatk4_reblockgvcf_dbsnp {
+    
+    input = [
+        [ id:'test', single_end:false ], // meta map
+        file(params.test_data['homo_sapiens']['illumina']['test_haplotc_cnn_vcf_gz'], checkIfExists: true),
+        file(params.test_data['homo_sapiens']['illumina']['test_haplotc_cnn_vcf_gz_tbi'], checkIfExists: true),
+        []
+    ]
+
+    fasta       = file(params.test_data['homo_sapiens']['genome']['genome_fasta'], checkIfExists: true)
+    fasta_index = file(params.test_data['homo_sapiens']['genome']['genome_fasta_fai'], checkIfExists: true)
+    dict        = file(params.test_data['homo_sapiens']['genome']['genome_dict'], checkIfExists: true)
+    dbsnp       = file(params.test_data['homo_sapiens']['genome']['dbsnp_146_hg38_vcf_gz'], checkIfExists: true)
+    dbsnp_tbi   = file(params.test_data['homo_sapiens']['genome']['dbsnp_146_hg38_vcf_gz_tbi'], checkIfExists: true)
+
+    GATK4_REBLOCKGVCF ( input, fasta, fasta_index, dict, dbsnp, dbsnp_tbi )
+}

tests/modules/gatk4/reblockgvcf/nextflow.config

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

tests/modules/gatk4/reblockgvcf/test.yml

@@ -0,0 +1,26 @@
+- name: gatk4 reblockgvcf test_gatk4_reblockgvcf
+  command: nextflow run ./tests/modules/gatk4/reblockgvcf -entry test_gatk4_reblockgvcf -c ./tests/config/nextflow.config  -c ./tests/modules/gatk4/reblockgvcf/nextflow.config
+  tags:
+    - gatk4/reblockgvcf
+    - gatk4
+  files:
+    - path: output/gatk4/test.rb.g.vcf.gz
+    - path: output/gatk4/test.rb.g.vcf.gz.tbi
+
+- name: gatk4 reblockgvcf test_gatk4_reblockgvcf_intervals
+  command: nextflow run ./tests/modules/gatk4/reblockgvcf -entry test_gatk4_reblockgvcf_intervals -c ./tests/config/nextflow.config  -c ./tests/modules/gatk4/reblockgvcf/nextflow.config
+  tags:
+    - gatk4/reblockgvcf
+    - gatk4
+  files:
+    - path: output/gatk4/test.rb.g.vcf.gz
+    - path: output/gatk4/test.rb.g.vcf.gz.tbi
+
+- name: gatk4 reblockgvcf test_gatk4_reblockgvcf_dbsnp
+  command: nextflow run ./tests/modules/gatk4/reblockgvcf -entry test_gatk4_reblockgvcf_dbsnp -c ./tests/config/nextflow.config  -c ./tests/modules/gatk4/reblockgvcf/nextflow.config
+  tags:
+    - gatk4/reblockgvcf
+    - gatk4
+  files:
+    - path: output/gatk4/test.rb.g.vcf.gz
+    - path: output/gatk4/test.rb.g.vcf.gz.tbi