add stub and the option to compress output for vep (#2111)

* add stub

* create tab and json files

* add option to get compressed outputs

modules/ensemblvep/main.nf

@@ -17,11 +17,14 @@ process ENSEMBLVEP {
     path  extra_files
 
     output:
-    tuple val(meta), path("*.ann.vcf")  , optional:true, emit: vcf
-    tuple val(meta), path("*.ann.tab")  , optional:true, emit: tab
-    tuple val(meta), path("*.ann.json") , optional:true, emit: json
-    path "*.summary.html"               , emit: report
-    path "versions.yml"                 , emit: versions
+    tuple val(meta), path("*.ann.vcf")     , optional:true, emit: vcf
+    tuple val(meta), path("*.ann.tab")     , optional:true, emit: tab
+    tuple val(meta), path("*.ann.json")    , optional:true, emit: json
+    tuple val(meta), path("*.ann.vcf.gz")  , optional:true, emit: vcf_gz
+    tuple val(meta), path("*.ann.tab.gz")  , optional:true, emit: tab_gz
+    tuple val(meta), path("*.ann.json.gz") , optional:true, emit: json_gz
+    path "*.summary.html"                  , emit: report
+    path "versions.yml"                    , emit: versions
 
     when:
     task.ext.when == null || task.ext.when
@@ -29,6 +32,7 @@ process ENSEMBLVEP {
     script:
     def args = task.ext.args ?: ''
     def file_extension = args.contains("--vcf") ? 'vcf' : args.contains("--json")? 'json' : args.contains("--tab")? 'tab' : 'vcf'
+    def compress_out = args.contains("--compress_output") ? '.gz' : ''
     def prefix = task.ext.prefix ?: "${meta.id}"
     def dir_cache = cache ? "\${PWD}/${cache}" : "/.vep"
     def reference = fasta ? "--fasta $fasta" : ""
@@ -36,7 +40,7 @@ process ENSEMBLVEP {
     """
     vep \\
         -i $vcf \\
-        -o ${prefix}.ann.${file_extension} \\
+        -o ${prefix}.ann.${file_extension}${compress_out} \\
         $args \\
         $reference \\
         --assembly $genome \\
@@ -48,6 +52,23 @@ process ENSEMBLVEP {
         --stats_file ${prefix}.summary.html \\
 
 
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        ensemblvep: \$( echo \$(vep --help 2>&1) | sed 's/^.*Versions:.*ensembl-vep : //;s/ .*\$//')
+    END_VERSIONS
+    """
+
+    stub:
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    touch ${prefix}.ann.vcf
+    touch ${prefix}.ann.tab
+    touch ${prefix}.ann.json
+    touch ${prefix}.ann.vcf.gz
+    touch ${prefix}.ann.tab.gz
+    touch ${prefix}.ann.json.gz
+    touch ${prefix}.summary.html
+
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":
         ensemblvep: \$( echo \$(vep --help 2>&1) | sed 's/^.*Versions:.*ensembl-vep : //;s/ .*\$//')

tests/modules/ensemblvep/main.nf

@@ -7,6 +7,8 @@ include { ENSEMBLVEP } from '../../../modules/ensemblvep/main.nf'
 include { ENSEMBLVEP as ENSEMBLVEP_JSON } from '../../../modules/ensemblvep/main.nf'
 include { ENSEMBLVEP as ENSEMBLVEP_TAB } from '../../../modules/ensemblvep/main.nf'
 include { ENSEMBLVEP as ENSEMBLVEP_VCF } from '../../../modules/ensemblvep/main.nf'
+include { ENSEMBLVEP as ENSEMBLVEP_VCF_BGZIP } from '../../../modules/ensemblvep/main.nf'
+include { ENSEMBLVEP as ENSEMBLVEP_VCF_GZIP  } from '../../../modules/ensemblvep/main.nf'
 
 workflow test_ensemblvep_fasta_json {
     input = [
@@ -41,6 +43,28 @@ workflow test_ensemblvep_fasta_vcf {
     ENSEMBLVEP_VCF ( input, "WBcel235", "caenorhabditis_elegans", "106", [], fasta, [] )
 }
 
+workflow test_ensemblvep_fasta_vcf_bgzip {
+    input = [
+        [ id:'test' ], // meta map
+        file(params.test_data['sarscov2']['illumina']['test_vcf'], checkIfExists: true)
+    ]
+
+    fasta = file(params.test_data['sarscov2']['genome']['genome_fasta'], checkIfExists: true)
+
+    ENSEMBLVEP_VCF_BGZIP ( input, "WBcel235", "caenorhabditis_elegans", "106", [], fasta, [] )
+}
+
+workflow test_ensemblvep_fasta_vcf_gzip {
+    input = [
+        [ id:'test' ], // meta map
+        file(params.test_data['sarscov2']['illumina']['test_vcf'], checkIfExists: true)
+    ]
+
+    fasta = file(params.test_data['sarscov2']['genome']['genome_fasta'], checkIfExists: true)
+
+    ENSEMBLVEP_VCF_GZIP ( input, "WBcel235", "caenorhabditis_elegans", "106", [], fasta, [] )
+}
+
 workflow test_ensemblvep_fasta {
     input = [
         [ id:'test' ], // meta map

tests/modules/ensemblvep/nextflow.config

@@ -20,4 +20,14 @@ process {
         container = 'nfcore/vep:106.1.WBcel235'
         ext.args = '--vcf'
     }
+
+    withName: ENSEMBLVEP_VCF_BGZIP {
+        container = 'nfcore/vep:106.1.WBcel235'
+        ext.args = '--vcf --compress_output bgzip'
+    }
+
+    withName: ENSEMBLVEP_VCF_GZIP {
+        container = 'nfcore/vep:106.1.WBcel235'
+        ext.args = '--vcf --compress_output gzip'
+    }
 }

tests/modules/ensemblvep/test.yml

@@ -22,6 +22,22 @@
     - path: output/ensemblvep/test.ann.vcf
     - path: output/ensemblvep/test.summary.html
 
+- name: ensemblvep test_ensemblvep_fasta_vcf_bgzip
+  command: nextflow run ./tests/modules/ensemblvep -entry test_ensemblvep_fasta_vcf_bgzip -c ./tests/config/nextflow.config  -c ./tests/modules/ensemblvep/nextflow.config
+  tags:
+    - ensemblvep
+  files:
+    - path: output/ensemblvep/test.ann.vcf.gz
+    - path: output/ensemblvep/test.summary.html
+
+- name: ensemblvep test_ensemblvep_fasta_vcf_gzip
+  command: nextflow run ./tests/modules/ensemblvep -entry test_ensemblvep_fasta_vcf_gzip -c ./tests/config/nextflow.config  -c ./tests/modules/ensemblvep/nextflow.config
+  tags:
+    - ensemblvep
+  files:
+    - path: output/ensemblvep/test.ann.vcf.gz
+    - path: output/ensemblvep/test.summary.html
+
 - name: ensemblvep test_ensemblvep_fasta
   command: nextflow run ./tests/modules/ensemblvep -entry test_ensemblvep_fasta -c ./tests/config/nextflow.config  -c ./tests/modules/ensemblvep/nextflow.config
   tags:
@@ -37,3 +53,11 @@
   files:
     - path: output/ensemblvep/test.ann.vcf
     - path: output/ensemblvep/test.summary.html
+
+- name: ensemblvep test_ensemblvep_fasta_vcf_stub
+  command: nextflow run ./tests/modules/ensemblvep -entry test_ensemblvep_fasta_vcf -c ./tests/config/nextflow.config  -c ./tests/modules/ensemblvep/nextflow.config -stub
+  tags:
+    - ensemblvep
+  files:
+    - path: output/ensemblvep/test.ann.vcf
+    - path: output/ensemblvep/test.summary.html