module: deepvariant

modules/deepvariant/functions.nf

@@ -0,0 +1,68 @@
+//
+//  Utility functions used in nf-core DSL2 module files
+//
+
+//
+// Extract name of software tool from process name using $task.process
+//
+def getSoftwareName(task_process) {
+    return task_process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()
+}
+
+//
+// Function to initialise default values and to generate a Groovy Map of available options for nf-core modules
+//
+def initOptions(Map args) {
+    def Map options = [:]
+    options.args            = args.args ?: ''
+    options.args2           = args.args2 ?: ''
+    options.args3           = args.args3 ?: ''
+    options.publish_by_meta = args.publish_by_meta ?: []
+    options.publish_dir     = args.publish_dir ?: ''
+    options.publish_files   = args.publish_files
+    options.suffix          = args.suffix ?: ''
+    return options
+}
+
+//
+// Tidy up and join elements of a list to return a path string
+//
+def getPathFromList(path_list) {
+    def paths = path_list.findAll { item -> !item?.trim().isEmpty() }      // Remove empty entries
+    paths     = paths.collect { it.trim().replaceAll("^[/]+|[/]+\$", "") } // Trim whitespace and trailing slashes
+    return paths.join('/')
+}
+
+//
+// Function to save/publish module results
+//
+def saveFiles(Map args) {
+    if (!args.filename.endsWith('.version.txt')) {
+        def ioptions  = initOptions(args.options)
+        def path_list = [ ioptions.publish_dir ?: args.publish_dir ]
+        if (ioptions.publish_by_meta) {
+            def key_list = ioptions.publish_by_meta instanceof List ? ioptions.publish_by_meta : args.publish_by_meta
+            for (key in key_list) {
+                if (args.meta && key instanceof String) {
+                    def path = key
+                    if (args.meta.containsKey(key)) {
+                        path = args.meta[key] instanceof Boolean ? "${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/deepvariant/main.nf

@@ -0,0 +1,58 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName } from './functions'
+
+// TODO nf-core: A module file SHOULD only define input and output files as command-line parameters.
+//               All other parameters MUST be provided as a string i.e. "options.args"
+//               where "params.options" is a Groovy Map that MUST be provided via the addParams section of the including workflow.
+//               Any parameters that need to be evaluated in the context of a particular sample
+//               e.g. single-end/paired-end data MUST also be defined and evaluated appropriately.
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process DEEPVARIANT {
+    tag "$meta.id"
+    label 'process_high'
+    publishDir "${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::deepvariant=1.1.0" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/deepvariant:1.1.0--py36hf3e76ba_2"
+    } else {
+        // TODO update the bioconda container to work with run_deepvariant.sh script
+        // container "quay.io/biocontainers/deepvariant:1.1.0--py36hf3e76ba_2"
+        container "google/deepvariant:1.1.0"
+    }
+
+    input:
+    // TODO nf-core: Where applicable all sample-specific information e.g. "id", "single_end", "read_group"
+    //               MUST be provided as an input via a Groovy Map called "meta".
+    //               This information may not be required in some instances e.g. indexing reference genome files:
+    //               https://github.com/nf-core/modules/blob/master/software/bwa/index/main.nf
+    tuple val(meta), path(bam), path(bai)
+    path fasta
+    path fai
+
+    output:
+    tuple val(meta), path("*.vcf.gz"),  emit: vcf
+    tuple val(meta), path("*g.vcf.gz"),  emit: gvcf
+    path "*.version.txt"          , emit: version
+
+    script:
+    def software = getSoftwareName(task.process)
+    def prefix   = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    """
+    /opt/deepvariant/bin/run_deepvariant \\
+        --ref=${fasta} \\
+        --reads=${bam} \\
+        --output_vcf=${prefix}.vcf.gz \\
+        --output_gvcf=${prefix}.g.vcf.gz \\
+        --num_shards=${task.cpus} \\
+        ${options.args}
+
+    echo \$(/opt/deepvariant/bin/run_deepvariant --version)  > ${software}.version.txt
+    """
+
+}

modules/deepvariant/meta.yml

@@ -0,0 +1,47 @@
+name: deepvariant
+## TODO nf-core: Add a description of the module and list keywords
+description: write your description here
+keywords:
+  - sort
+tools:
+  - deepvariant:
+      ## TODO nf-core: Add a description and other details for the software below
+      description: DeepVariant is an analysis pipeline that uses a deep neural network to call genetic variants from next-generation DNA sequencing data
+      homepage: None
+      documentation: None
+      tool_dev_url: None
+      doi: ""
+      licence: ['MIT']
+
+## TODO nf-core: Add a description of all of the variables used as input
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  ## TODO nf-core: Delete / customise this example input
+  - bam:
+      type: file
+      description: BAM/CRAM/SAM file
+      pattern: "*.{bam,cram,sam}"
+
+## TODO nf-core: Add a description of all of the variables used as output
+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: "*.{version.txt}"
+  ## TODO nf-core: Delete / customise this example output
+  - bam:
+      type: file
+      description: Sorted BAM/CRAM/SAM file
+      pattern: "*.{bam,cram,sam}"
+
+authors:
+  - "@abhi18av"

tests/config/pytest_modules.yml

@@ -242,6 +242,10 @@ deeptools/plotprofile:
   - modules/deeptools/plotprofile/**
   - tests/modules/deeptools/plotprofile/**
 
+deepvariant:
+  - modules/deepvariant/**
+  - tests/modules/deepvariant/**
+
 delly/call:
   - modules/delly/call/**
   - tests/modules/delly/call/**

tests/modules/deepvariant/main.nf

@@ -0,0 +1,13 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { DEEPVARIANT } from '../../../modules/deepvariant/main.nf' addParams( options: [:] )
+
+workflow test_deepvariant {
+
+    input = [ [ id:'test', single_end:false ], // meta map
+              file(params.test_data['sarscov2']['illumina']['test_paired_end_bam'], checkIfExists: true) ]
+
+    DEEPVARIANT ( input )
+}

tests/modules/deepvariant/test.yml

@@ -0,0 +1,9 @@
+## TODO nf-core: Please run the following command to build this file:
+#                nf-core modules create-test-yml deepvariant
+- name: deepvariant
+  command: nextflow run ./tests/modules/deepvariant -entry test_deepvariant -c tests/config/nextflow.config
+  tags:
+    - deepvariant
+  files:
+    - path: output/deepvariant/test.bam
+      md5sum: e667c7caad0bc4b7ac383fd023c654fc