Add new vcflib/vcfuniq module (#1208)

* Add new vcflib/vcfuniq module

* Rename module from vcfutils to vcfuniq

* Fix tests

* Compress output by default

* Fix ECLint

modules/vcflib/vcfuniq/main.nf

@@ -0,0 +1,32 @@
+def VERSION = '1.0.2' // Version information not provided by tool on CLI
+
+process VCFLIB_VCFUNIQ {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda (params.enable_conda ? "bioconda::vcflib=1.0.2" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/vcflib:1.0.2--h3198e80_5':
+        'quay.io/biocontainers/vcflib:1.0.2--h3198e80_5' }"
+
+    input:
+    tuple val(meta), path(vcf), path(tbi)
+
+    output:
+    tuple val(meta), path("*.gz"), emit: vcf
+    path "versions.yml"          , emit: versions
+
+    script:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    vcfuniq \\
+        $vcf \\
+        | bgzip -c $args > ${prefix}.vcf.gz
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        vcflib: $VERSION
+    END_VERSIONS
+    """
+}

modules/vcflib/vcfuniq/meta.yml

@@ -0,0 +1,46 @@
+name: vcflib_vcfuniq
+description: List unique genotypes. Like GNU uniq, but for VCF records. Remove records which have the same position, ref, and alt as the previous record.
+keywords:
+  - vcf
+  - uniq
+  - deduplicate
+tools:
+  - vcflib:
+      description: Command-line tools for manipulating VCF files
+      homepage: https://github.com/vcflib/vcflib
+      documentation: https://github.com/vcflib/vcflib#USAGE
+      doi: "https://doi.org/10.1101/2021.05.21.445151"
+      licence: ['MIT']
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - vcf:
+      type: file
+      description: Compressed VCF file
+      pattern: "*.vcf.gz"
+  - tbi:
+      type: file
+      description: Index of VCF file
+      pattern: "*.vcf.gz.tbi"
+
+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: Compressed VCF file
+      pattern: "*.vcf.gz"
+
+authors:
+  - "@drpatelh"

tests/config/pytest_modules.yml

@@ -1485,6 +1485,10 @@ variantbam:
   - modules/variantbam/**
   - tests/modules/variantbam/**
 
+vcflib/vcfuniq:
+  - modules/vcflib/vcfuniq/**
+  - tests/modules/vcflib/vcfuniq/**
+
 vcftools:
   - modules/vcftools/**
   - tests/modules/vcftools/**

tests/modules/vcflib/vcfuniq/main.nf

@@ -0,0 +1,16 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { VCFLIB_VCFUNIQ } from '../../../../modules/vcflib/vcfuniq/main.nf'
+
+workflow test_vcflib_vcfuniq {
+    
+    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)
+    ]
+
+    VCFLIB_VCFUNIQ ( input )
+}

tests/modules/vcflib/vcfuniq/nextflow.config

@@ -0,0 +1,9 @@
+process {
+
+    publishDir = { "${params.outdir}/${task.process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()}" }
+    
+    withName: VCFLIB_VCFUNIQ {
+        ext.prefix = { "${meta.id}.uniq" }
+    }
+
+}

tests/modules/vcflib/vcfuniq/test.yml

@@ -0,0 +1,8 @@
+- name: vcflib vcfuniq
+  command: nextflow run ./tests/modules/vcflib/vcfuniq -entry test_vcflib_vcfuniq -c ./tests/config/nextflow.config -c ./tests/modules/vcflib/vcfuniq/nextflow.config
+  tags:
+    - vcflib
+    - vcflib/vcfuniq
+  files:
+    - path: output/vcflib/test.uniq.vcf.gz
+      md5sum: fc178eb342a91dc0d1d568601ad8f8e2