Add gatk4/baserecalibrator (#327)

* Add gatk4 BaseRecalibrator

* Solve format in main.nf

* Correct functions

* Add pytest

software/gatk4/baserecalibrator/functions.nf

@@ -0,0 +1,60 @@
+/*
+ * -----------------------------------------------------
+ *  Utility functions used in nf-core DSL2 module files
+ * -----------------------------------------------------
+ */
+
+/*
+ * Extract name of software tool from process name using $task.process
+ */
+def getSoftwareName(task_process) {
+    return task_process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()
+}
+
+/*
+ * Function to initialise default values and to generate a Groovy Map of available options for nf-core modules
+ */
+def initOptions(Map args) {
+    def Map options = [:]
+    options.args          = args.args ?: ''
+    options.args2         = args.args2 ?: ''
+    options.args3         = args.args3 ?: ''
+    options.publish_by_id = args.publish_by_id ?: false
+    options.publish_dir   = args.publish_dir ?: ''
+    options.publish_files = args.publish_files
+    options.suffix        = args.suffix ?: ''
+    return options
+}
+
+/*
+ * Tidy up and join elements of a list to return a path string
+ */
+def getPathFromList(path_list) {
+    def paths = path_list.findAll { item -> !item?.trim().isEmpty() }  // Remove empty entries
+    paths = paths.collect { it.trim().replaceAll("^[/]+|[/]+\$", "") } // Trim whitespace and trailing slashes
+    return paths.join('/')
+}
+
+/*
+ * Function to save/publish module results
+ */
+def saveFiles(Map args) {
+    if (!args.filename.endsWith('.version.txt')) {
+        def ioptions = initOptions(args.options)
+        def path_list = [ ioptions.publish_dir ?: args.publish_dir ]
+        if (ioptions.publish_by_id) {
+            path_list.add(args.publish_id)
+        }
+        if (ioptions.publish_files instanceof Map) {
+            for (ext in ioptions.publish_files) {
+                if (args.filename.endsWith(ext.key)) {
+                    def ext_list = path_list.collect()
+                    ext_list.add(ext.value)
+                    return "${getPathFromList(ext_list)}/$args.filename"
+                }
+            }
+        } else if (ioptions.publish_files == null) {
+            return "${getPathFromList(path_list)}/$args.filename"
+        }
+    }
+}

software/gatk4/baserecalibrator/main.nf

@@ -0,0 +1,50 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process GATK4_BASERECALIBRATOR {
+    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), publish_id:meta.id) }
+
+    conda (params.enable_conda ? "bioconda::gatk4=4.2.0.0" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/gatk4:4.2.0.0--0"
+    } else {
+        container "quay.io/biocontainers/gatk4:4.2.0.0--0"
+    }
+
+    input:
+    tuple val(meta), path(bam)
+    path fasta
+    path fastaidx
+    path dict
+    path intervalsBed
+    path knownSites
+    path knownSites_tbi
+
+    output:
+    tuple val(meta), path("*.table"), emit: table
+    path "*.version.txt"          , emit: version
+
+    script:
+    def software = getSoftwareName(task.process)
+    def prefix   = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    def intervalsCommand = intervalsBed ? "-L ${intervalsBed}" : ""
+    def sitesCommand = knownSites.collect{"--known-sites ${it}"}.join(' ')
+    """
+    gatk BaseRecalibrator  \
+        -R $fasta \
+        -I $bam \
+        $sitesCommand \
+        $intervalsCommand \
+        $options.args \
+        -O ${prefix}.table
+
+    gatk --version | grep Picard | sed "s/Picard Version: //g" > ${software}.version.txt
+    """
+}

software/gatk4/baserecalibrator/meta.yml

@@ -0,0 +1,58 @@
+name: gatk4_baserecalibrator
+description: Generate recalibration table for Base Quality Score Recalibration (BQSR)
+keywords:
+  - sort
+tools:
+  - gatk4:
+    description: |
+        Developed in the Data Sciences Platform at the Broad Institute, the toolkit offers a wide variety of tools
+        with a primary focus on variant discovery and genotyping. Its powerful processing engine
+        and high-performance computing features make it capable of taking on projects of any size.
+    homepage: https://gatk.broadinstitute.org/hc/en-us
+    documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
+    doi: 10.1158/1538-7445.AM2017-3590
+
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - bam:
+      type: file
+      description: BAM file from alignment
+      pattern: "*.{bam}"
+  - fasta:
+      type: file
+      description: The reference fasta file
+  - fastaidx:
+      type: file
+      description: Index of reference fasta file
+  - dict:
+      type: file
+      description: GATK sequence dictionary
+  - intervalsBed:
+      type: file
+      description: Bed file with the genomic regions included in the library (optional)
+  - knownSites:
+      type: file
+      description: Bed file with the genomic regions included in the library (optional)
+
+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}"
+  - table:
+      type: file
+      description: Recalibration table from BaseRecalibrator
+      pattern: "*.{table}"
+
+authors:
+  - "@yocra3"

tests/config/pytest_software.yml

@@ -182,6 +182,10 @@ gatk4_applybqsr:
   - software/gatk4/applybqsr/**
   - tests/software/gatk4/applybqsr/**
 
+gatk4_baserecalibrator:
+  - software/gatk4/baserecalibrator/**
+  - tests/software/gatk4/baserecalibrator/**
+
 gatk4_bedtointervallist:
   - software/gatk4/bedtointervallist/**
   - tests/software/gatk4/bedtointervallist/**

tests/software/gatk4/baserecalibrator/main.nf

@@ -0,0 +1,50 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { GATK4_BASERECALIBRATOR } from '../../../../software/gatk4/baserecalibrator/main.nf' addParams( options: [:] )
+
+workflow test_gatk4_baserecalibrator {
+
+  def input = []
+  input = [ [ id:'test' ], // meta map
+            file("${launchDir}/tests/data/genomics/sarscov2/bam/test_paired_end.sorted.bam", checkIfExists: true) ]
+  fasta = file("${launchDir}/tests/data/genomics/sarscov2/fasta/test_genome.fasta", checkIfExists: true)
+  fastaidx = file("${launchDir}/tests/data/genomics/sarscov2/fasta/test_genome.fasta.fai", checkIfExists: true)
+  dict = file("${launchDir}/tests/data/genomics/sarscov2/fasta/test_genome.dict", checkIfExists: true)
+
+  sites = file("${launchDir}/tests/data/genomics/sarscov2/vcf/test.vcf.gz", checkIfExists: true)
+  sites_tbi = file("${launchDir}/tests/data/genomics/sarscov2/vcf/test.vcf.gz.tbi", checkIfExists: true)
+  GATK4_BASERECALIBRATOR ( input, fasta, fastaidx, dict, [], sites, sites_tbi )
+}
+
+workflow test_gatk4_baserecalibrator_intervals {
+
+  def input = []
+  input = [ [ id:'test' ], // meta map
+            file("${launchDir}/tests/data/genomics/sarscov2/bam/test_paired_end.sorted.bam", checkIfExists: true) ]
+  fasta = file("${launchDir}/tests/data/genomics/sarscov2/fasta/test_genome.fasta", checkIfExists: true)
+  fastaidx = file("${launchDir}/tests/data/genomics/sarscov2/fasta/test_genome.fasta.fai", checkIfExists: true)
+  dict = file("${launchDir}/tests/data/genomics/sarscov2/fasta/test_genome.dict", checkIfExists: true)
+  intervals = file("${launchDir}/tests/data/genomics/sarscov2/bed/test.bed", checkIfExists: true)
+  sites = file("${launchDir}/tests/data/genomics/sarscov2/vcf/test.vcf.gz", checkIfExists: true)
+  sites_tbi = file("${launchDir}/tests/data/genomics/sarscov2/vcf/test.vcf.gz.tbi", checkIfExists: true)
+  GATK4_BASERECALIBRATOR ( input, fasta, fastaidx, dict, intervals, sites, sites_tbi )
+}
+
+
+workflow test_gatk4_baserecalibrator_multiple_sites {
+
+  def input = []
+  input = [ [ id:'test' ], // meta map
+            file("${launchDir}/tests/data/genomics/sarscov2/bam/test_paired_end.sorted.bam", checkIfExists: true) ]
+  fasta = file("${launchDir}/tests/data/genomics/sarscov2/fasta/test_genome.fasta", checkIfExists: true)
+  fastaidx = file("${launchDir}/tests/data/genomics/sarscov2/fasta/test_genome.fasta.fai", checkIfExists: true)
+  dict = file("${launchDir}/tests/data/genomics/sarscov2/fasta/test_genome.dict", checkIfExists: true)
+
+  sites = [file("${launchDir}/tests/data/genomics/sarscov2/vcf/test.vcf.gz", checkIfExists: true),
+           file("${launchDir}/tests/data/genomics/sarscov2/vcf/test2.vcf.gz", checkIfExists: true)]
+  sites_tbi = [file("${launchDir}/tests/data/genomics/sarscov2/vcf/test.vcf.gz.tbi", checkIfExists: true),
+              file("${launchDir}/tests/data/genomics/sarscov2/vcf/test2.vcf.gz.tbi", checkIfExists: true)]
+  GATK4_BASERECALIBRATOR ( input, fasta, fastaidx, dict, [], sites, sites_tbi )
+}

tests/software/gatk4/baserecalibrator/test.yml

@@ -0,0 +1,28 @@
+- name: gatk4 baserecalibrator test_gatk4_baserecalibrator
+  command: nextflow run tests/software/gatk4/baserecalibrator -entry test_gatk4_baserecalibrator -c tests/config/nextflow.config
+  tags:
+    - gatk4
+    - gatk4_baserecalibrator
+  files:
+    - path: output/gatk4/test.table
+      md5sum: e2e43abdc0c943c1a54dae816d0b9ea7
+
+- name: gatk4 baserecalibrator test_gatk4_baserecalibrator_intervals
+  command: nextflow run tests/software/gatk4/baserecalibrator -entry test_gatk4_baserecalibrator_intervals -c tests/config/nextflow.config
+  tags:
+    - gatk4
+    - gatk4_baserecalibrator
+    - gatk4_baserecalibrator_intervals
+  files:
+    - path: output/gatk4/test.table
+      md5sum: 9ecb5f00a2229291705addc09c0ec231
+
+- name: gatk4 baserecalibrator test_gatk4_baserecalibrator_multiple_sites
+  command: nextflow run tests/software/gatk4/baserecalibrator -entry test_gatk4_baserecalibrator_multiple_sites -c tests/config/nextflow.config
+  tags:
+    - gatk4_baserecalibrator_multiple_sites
+    - gatk4
+    - gatk4_baserecalibrator
+  files:
+    - path: output/gatk4/test.table
+      md5sum: e2e43abdc0c943c1a54dae816d0b9ea7