Add bcftools/index (#812)

* feat: add bcftools index

* Extend tests to also test gen for TBI

* Update meta.yml

* Update meta.yml

modules/bcftools/index/functions.nf

@@ -0,0 +1,78 @@
+//
+//  Utility functions used in nf-core DSL2 module files
+//
+
+//
+// Extract name of software tool from process name using $task.process
+//
+def getSoftwareName(task_process) {
+    return task_process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()
+}
+
+//
+// Extract name of module from process name using $task.process
+//
+def getProcessName(task_process) {
+    return task_process.tokenize(':')[-1]
+}
+
+//
+// Function to initialise default values and to generate a Groovy Map of available options for nf-core modules
+//
+def initOptions(Map args) {
+    def Map options = [:]
+    options.args            = args.args ?: ''
+    options.args2           = args.args2 ?: ''
+    options.args3           = args.args3 ?: ''
+    options.publish_by_meta = args.publish_by_meta ?: []
+    options.publish_dir     = args.publish_dir ?: ''
+    options.publish_files   = args.publish_files
+    options.suffix          = args.suffix ?: ''
+    return options
+}
+
+//
+// Tidy up and join elements of a list to return a path string
+//
+def getPathFromList(path_list) {
+    def paths = path_list.findAll { item -> !item?.trim().isEmpty() }      // Remove empty entries
+    paths     = paths.collect { it.trim().replaceAll("^[/]+|[/]+\$", "") } // Trim whitespace and trailing slashes
+    return paths.join('/')
+}
+
+//
+// Function to save/publish module results
+//
+def saveFiles(Map args) {
+    def ioptions  = initOptions(args.options)
+    def path_list = [ ioptions.publish_dir ?: args.publish_dir ]
+
+    // Do not publish versions.yml unless running from pytest workflow
+    if (args.filename.equals('versions.yml') && !System.getenv("NF_CORE_MODULES_TEST")) {
+        return null
+    }
+    if (ioptions.publish_by_meta) {
+        def key_list = ioptions.publish_by_meta instanceof List ? ioptions.publish_by_meta : args.publish_by_meta
+        for (key in key_list) {
+            if (args.meta && key instanceof String) {
+                def path = key
+                if (args.meta.containsKey(key)) {
+                    path = args.meta[key] instanceof Boolean ? "${key}_${args.meta[key]}".toString() : args.meta[key]
+                }
+                path = path instanceof String ? path : ''
+                path_list.add(path)
+            }
+        }
+    }
+    if (ioptions.publish_files instanceof Map) {
+        for (ext in ioptions.publish_files) {
+            if (args.filename.endsWith(ext.key)) {
+                def ext_list = path_list.collect()
+                ext_list.add(ext.value)
+                return "${getPathFromList(ext_list)}/$args.filename"
+            }
+        }
+    } else if (ioptions.publish_files == null) {
+        return "${getPathFromList(path_list)}/$args.filename"
+    }
+}

modules/bcftools/index/main.nf

@@ -0,0 +1,44 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName; getProcessName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process BCFTOOLS_INDEX {
+    tag "$meta.id"
+    label 'process_low'
+    publishDir "${params.outdir}",
+        mode: params.publish_dir_mode,
+        saveAs: { filename -> saveFiles(filename:filename, options:params.options, publish_dir:getSoftwareName(task.process), meta:meta, publish_by_meta:['id']) }
+
+    conda (params.enable_conda ? 'bioconda::bcftools=1.13' : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/bcftools:1.13--h3a49de5_0"
+    } else {
+        container "quay.io/biocontainers/bcftools:1.13--h3a49de5_0"
+    }
+
+    input:
+    tuple val(meta), path(vcf)
+
+    output:
+    tuple val(meta), path("*.csi"), optional:true, emit: csi
+    tuple val(meta), path("*.tbi"), optional:true, emit: tbi
+    path "versions.yml"          , emit: version
+
+    script:
+    def prefix   = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+
+    """
+    bcftools \\
+        index \\
+        $options.args \\
+        --threads $task.cpus \\
+        $vcf
+
+    cat <<-END_VERSIONS > versions.yml
+    ${getProcessName(task.process)}:
+        ${getSoftwareName(task.process)}: \$(bcftools --version 2>&1 | head -n1 | sed 's/^.*bcftools //; s/ .*\$//')
+    END_VERSIONS
+    """
+}

modules/bcftools/index/meta.yml

@@ -0,0 +1,49 @@
+name: bcftools_index
+description: Index VCF tools
+keywords:
+  - vcf
+  - index
+  - bcftools
+  - csi
+  - tbi
+tools:
+  - bcftools:
+      description: BCFtools is a set of utilities that manipulate variant calls in the Variant Call Format (VCF) and its binary counterpart BCF. All commands work transparently with both VCFs and BCFs, both uncompressed and BGZF-compressed.  Most commands accept VCF, bgzipped VCF and BCF with filetype detected automatically even when streaming from a pipe. Indexed VCF and BCF will work in all situations. Un-indexed VCF and BCF and streams will work in most, but not all situations.
+      homepage: https://samtools.github.io/bcftools/
+      documentation: https://samtools.github.io/bcftools/howtos/index.html
+      tool_dev_url: https://github.com/samtools/bcftools
+      doi: "10.1093/gigascience/giab008"
+      licence: ['GPL']
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - VCF:
+      type: file
+      description: VCF file (optionally GZIPPED)
+      pattern: "*.{vcf,vcf.gz}"
+
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - version:
+      type: file
+      description: File containing software version
+      pattern: "versions.yml"
+  - csi:
+      type: file
+      description: Default VCF file index file
+      pattern: "*.csi"
+  - tbi:
+      type: file
+      description: Alternative VCF file index file for larger files (activated with -t parameter)
+      pattern: "*.tbi"
+
+authors:
+  - "@jfy133"

tests/config/pytest_modules.yml

@@ -70,6 +70,10 @@ bcftools/filter:
   - modules/bcftools/filter/**
   - tests/modules/bcftools/filter/**
 
+bcftools/index:
+  - modules/bcftools/index/**
+  - tests/modules/bcftools/index**
+
 bcftools/isec:
   - modules/bcftools/isec/**
   - tests/modules/bcftools/isec/**

tests/modules/bcftools/index/main.nf

@@ -0,0 +1,23 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { BCFTOOLS_INDEX as BCFTOOLS_INDEX_CSI } from '../../../../modules/bcftools/index/main.nf' addParams( options: [:] )
+include { BCFTOOLS_INDEX as BCFTOOLS_INDEX_TBI } from '../../../../modules/bcftools/index/main.nf' addParams( options: [args: '-t'] )
+
+
+workflow test_bcftools_index_csi {
+
+    input = [ [ id:'test', single_end:false ], // meta map
+              file(params.test_data['sarscov2']['illumina']['test_vcf_gz'], checkIfExists: true) ]
+
+    BCFTOOLS_INDEX_CSI ( input )
+}
+
+workflow test_bcftools_index_tbi {
+
+    input = [ [ id:'test', single_end:false ], // meta map
+              file(params.test_data['sarscov2']['illumina']['test_vcf_gz'], checkIfExists: true) ]
+
+    BCFTOOLS_INDEX_TBI ( input )
+}

tests/modules/bcftools/index/test.yml

@@ -0,0 +1,17 @@
+- name: bcftools index
+  command: nextflow run ./tests/modules/bcftools/index -entry test_bcftools_index_csi -c tests/config/nextflow.config
+  tags:
+    - bcftools
+    - bcftools/index
+  files:
+    - path: output/bcftools/test.vcf.gz.csi
+      md5sum: 5f930522d2b9dcdba2807b7da4dfa3fd
+
+- name: bcftools index tbi
+  command: nextflow run ./tests/modules/bcftools/index -entry test_bcftools_index_tbi -c tests/config/nextflow.config
+  tags:
+    - bcftools
+    - bcftools/index
+  files:
+    - path: output/bcftools/test.vcf.gz.tbi
+      md5sum: 36e11bf96ed0af4a92caa91a68612d64