Merge pull request #1758 from FriederikeHanssen/manta_convert

add manta conversion script

modules/manta/convertinversion/main.nf

@@ -0,0 +1,35 @@
+process MANTA_CONVERTINVERSION {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda (params.enable_conda ? "bioconda::manta=1.6.0 bioconda::samtools=1.15.1" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/mulled-v2-40295ae41112676b05b649e513fe7000675e9b84:0b4be2c719f99f44df34be7b447b287bb7f86e01-0':
+        'quay.io/biocontainers/mulled-v2-40295ae41112676b05b649e513fe7000675e9b84:0b4be2c719f99f44df34be7b447b287bb7f86e01-0' }"
+
+    input:
+    tuple val(meta), path(vcf)
+    path fasta
+
+    output:
+    tuple val(meta), path("*.vcf.gz")    , emit: vcf
+    tuple val(meta), path("*.vcf.gz.tbi"), emit: tbi
+    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}"
+    """
+    convertInversion.py \$(which samtools) $fasta $vcf | bgzip --threads $task.cpus > ${prefix}.vcf.gz
+    tabix ${prefix}.vcf.gz
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        manta: \$( configManta.py --version )
+        samtools: \$(echo \$(samtools --version 2>&1) | sed 's/^.*samtools //; s/Using.*\$//' ))
+    END_VERSIONS
+    """
+}

modules/manta/convertinversion/meta.yml

@@ -0,0 +1,45 @@
+name: "manta_convertinversion"
+description: Manta calls structural variants (SVs) and indels from mapped paired-end sequencing reads. This script reformats inversions into single inverted sequence junctions which was the format used in Manta versions <= 1.4.0.
+keywords:
+  - structural variants
+  - conversion
+tools:
+  - manta:
+      description: Structural variant and indel caller for mapped sequencing data
+      homepage: https://github.com/Illumina/manta
+      documentation: https://github.com/Illumina/manta/blob/v1.6.0/docs/userGuide/README.md
+      tool_dev_url: https://github.com/Illumina/manta
+      doi: "10.1093/bioinformatics/btv710"
+      licence: ["GPL v3"]
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - vcf:
+      type: file
+      description: VCF file produces by Manta
+      pattern: "*.vcf.gz"
+
+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"
+  - vcf:
+      type: file
+      description: VCF file with reformatted inversions
+      pattern: "*.vcf.gz"
+  - tbi:
+      type: file
+      description: TBI file produces by Manta
+      pattern: "*.vcf.gz.tbi"
+authors:
+  - "@FriederikeHanssen"

tests/config/pytest_modules.yml

@@ -1246,6 +1246,10 @@ maltextract:
   - modules/maltextract/**
   - tests/modules/maltextract/**
 
+manta/convertinversion:
+  - modules/manta/convertinversion/**
+  - tests/modules/manta/convertinversion/**
+
 manta/germline:
   - modules/manta/germline/**
   - tests/modules/manta/germline/**

tests/modules/manta/convertinversion/main.nf

@@ -0,0 +1,23 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { MANTA_CONVERTINVERSION } from '../../../../modules/manta/convertinversion/main.nf'
+include { MANTA_TUMORONLY        } from '../../../../modules/manta/tumoronly/main.nf'
+
+workflow test_manta_convertinversion {
+
+    input = [
+        [ id:'test'], // meta map
+        file(params.test_data['homo_sapiens']['illumina']['test2_paired_end_recalibrated_sorted_cram'], checkIfExists: true),
+        file(params.test_data['homo_sapiens']['illumina']['test2_paired_end_recalibrated_sorted_cram_crai'], checkIfExists: true),
+        [], []
+    ]
+
+    fasta   = file(params.test_data['homo_sapiens']['genome']['genome_21_fasta'], checkIfExists: true)
+    fai     = file(params.test_data['homo_sapiens']['genome']['genome_21_fasta_fai'], checkIfExists: true)
+
+    MANTA_TUMORONLY ( input, fasta, fai )
+
+    MANTA_CONVERTINVERSION ( MANTA_TUMORONLY.out.tumor_sv_vcf, fasta )
+}

tests/modules/manta/convertinversion/nextflow.config

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

tests/modules/manta/convertinversion/test.yml

@@ -0,0 +1,18 @@
+- name: manta convertinversion test_manta_convertinversion
+  command: nextflow run ./tests/modules/manta/convertinversion -entry test_manta_convertinversion -c ./tests/config/nextflow.config  -c ./tests/modules/manta/convertinversion/nextflow.config
+  tags:
+    - manta
+    - manta/convertinversion
+  files:
+    - path: output/manta/test.candidate_small_indels.vcf.gz
+    - path: output/manta/test.candidate_small_indels.vcf.gz.tbi
+      md5sum: 4cb176febbc8c26d717a6c6e67b9c905
+    - path: output/manta/test.candidate_sv.vcf.gz
+    - path: output/manta/test.candidate_sv.vcf.gz.tbi
+      md5sum: 4cb176febbc8c26d717a6c6e67b9c905
+    - path: output/manta/test.tumor_sv.vcf.gz
+    - path: output/manta/test.tumor_sv.vcf.gz.tbi
+      md5sum: 4cb176febbc8c26d717a6c6e67b9c905
+    - path: output/manta/test.vcf.gz
+    - path: output/manta/test.vcf.gz.tbi
+      md5sum: e7180bb953d2bd657c420a5f76a7164d