From 435ca4100ac07d1932cd0cb38226b7c4b2109fc1 Mon Sep 17 00:00:00 2001
From: Mei Wu <25568561+projectoriented@users.noreply.github.com>
Date: Thu, 20 Jan 2022 09:46:31 +0100
Subject: [PATCH] New module/vcfanno (#1204)

* boilerplate

* tests passed

* fixed regex

Co-authored-by: Chris Cheshire <chris.j.cheshire@gmail.com>
---
 modules/vcfanno/main.nf               | 34 ++++++++++++++++
 modules/vcfanno/meta.yml              | 56 +++++++++++++++++++++++++++
 tests/config/pytest_modules.yml       |  4 ++
 tests/modules/vcfanno/main.nf         | 19 +++++++++
 tests/modules/vcfanno/nextflow.config |  5 +++
 tests/modules/vcfanno/test.yml        |  9 +++++
 6 files changed, 127 insertions(+)
 create mode 100644 modules/vcfanno/main.nf
 create mode 100644 modules/vcfanno/meta.yml
 create mode 100644 tests/modules/vcfanno/main.nf
 create mode 100644 tests/modules/vcfanno/nextflow.config
 create mode 100644 tests/modules/vcfanno/test.yml

diff --git a/modules/vcfanno/main.nf b/modules/vcfanno/main.nf
new file mode 100644
index 00000000..421fc0fa
--- /dev/null
+++ b/modules/vcfanno/main.nf
@@ -0,0 +1,34 @@
+process VCFANNO {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda (params.enable_conda ? "bioconda::vcfanno=0.3.3" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/vcfanno:0.3.3--h9ee0642_0':
+        'quay.io/biocontainers/vcfanno:0.3.3--h9ee0642_0' }"
+
+    input:
+    tuple val(meta), path(vcf), path(tbi)
+    path vcfanno_config
+
+    output:
+    tuple val(meta), path("*.vcf"), emit: vcf
+    path "versions.yml"           , emit: versions
+
+    script:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    vcfanno \\
+        -p $task.cpus \\
+        $args \\
+        $vcfanno_config \\
+        $vcf \\
+        > ${prefix}_annotated.vcf
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        vcfanno: \$(echo \$(vcfanno 2>&1 | grep version | cut -f3 -d' ' ))
+    END_VERSIONS
+    """
+}
diff --git a/modules/vcfanno/meta.yml b/modules/vcfanno/meta.yml
new file mode 100644
index 00000000..bbd2ab16
--- /dev/null
+++ b/modules/vcfanno/meta.yml
@@ -0,0 +1,56 @@
+name: vcfanno
+description: quickly annotate your VCF with any number of INFO fields from any number of VCFs or BED files
+keywords:
+  - vcf
+  - bed
+  - annotate
+  - variant
+tools:
+  - vcfanno:
+      description: annotate a VCF with other VCFs/BEDs/tabixed files
+      homepage: None
+      documentation: https://github.com/brentp/vcfanno#vcfanno
+      tool_dev_url: https://github.com/brentp/vcfanno
+      doi: "10.1186/s13059-016-0973-5"
+      licence: ["MIT"]
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - vcf:
+      type: file
+      description: query VCF file
+      pattern: "*.{vcf.gz}"
+  - vcf.tbi:
+      type: file
+      description: query VCF file index
+      pattern: "*.{vcf.gz.tbi}"
+  - vcfanno_config:
+      type: file
+      description: |
+        A simple configuration file is used to specify both the source files
+        and the set of attributes (in the case of VCF)
+        or columns (in the case of BED or other tab-delimited formats)
+        that should be added to the query file.
+      pattern: "*.{toml}"
+
+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: Annotated VCF file
+      pattern: "*.{vcf}"
+
+authors:
+  - "@projectoriented"
diff --git a/tests/config/pytest_modules.yml b/tests/config/pytest_modules.yml
index fc062e58..967b7d9f 100644
--- a/tests/config/pytest_modules.yml
+++ b/tests/config/pytest_modules.yml
@@ -1493,6 +1493,10 @@ variantbam:
   - modules/variantbam/**
   - tests/modules/variantbam/**
 
+vcfanno:
+  - modules/vcfanno/**
+  - tests/modules/vcfanno/**
+
 vcflib/vcfuniq:
   - modules/vcflib/vcfuniq/**
   - tests/modules/vcflib/vcfuniq/**
diff --git a/tests/modules/vcfanno/main.nf b/tests/modules/vcfanno/main.nf
new file mode 100644
index 00000000..c46afa8f
--- /dev/null
+++ b/tests/modules/vcfanno/main.nf
@@ -0,0 +1,19 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { VCFANNO } from '../../../modules/vcfanno/main.nf'
+
+workflow test_vcfanno {
+    
+    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)
+    ]
+
+    toml = file("https://raw.githubusercontent.com/nf-core/test-datasets/8fbd9f99a2feb3f9e39cd3bcdc4a9176a5835673/data/delete_me/vcfanno.toml", 
+                checkIfExists: true)
+
+    VCFANNO ( input, toml )
+}
diff --git a/tests/modules/vcfanno/nextflow.config b/tests/modules/vcfanno/nextflow.config
new file mode 100644
index 00000000..50f50a7a
--- /dev/null
+++ b/tests/modules/vcfanno/nextflow.config
@@ -0,0 +1,5 @@
+process {
+
+    publishDir = { "${params.outdir}/${task.process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()}" }
+    
+}
\ No newline at end of file
diff --git a/tests/modules/vcfanno/test.yml b/tests/modules/vcfanno/test.yml
new file mode 100644
index 00000000..0b5394d2
--- /dev/null
+++ b/tests/modules/vcfanno/test.yml
@@ -0,0 +1,9 @@
+- name: vcfanno test_vcfanno
+  command: nextflow run tests/modules/vcfanno -entry test_vcfanno -c tests/config/nextflow.config
+  tags:
+    - vcfanno
+  files:
+    - path: output/vcfanno/test_annotated.vcf
+      md5sum: 34259cf6b0a4698a2917ad3554b50c0f
+    - path: output/vcfanno/versions.yml
+      md5sum: 62d13540503b22f04a2280c91942cb03