new module: samtools/fastq (#316)

* new module: samtools/fastq

* solve conflict: pytest_software.yml

* solve linting conflicts

* solved EditorConfig linting problem

software/samtools/fastq/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/samtools/fastq/main.nf

@@ -0,0 +1,40 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process SAMTOOLS_FASTQ {
+    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::samtools=1.12" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/samtools:1.12--hd5e65b6_0"
+    } else {
+        container "quay.io/biocontainers/samtools:1.12--hd5e65b6_0"
+    }
+
+    input:
+    tuple val(meta), path(bam)
+
+    output:
+    tuple val(meta), path("*.fastq"), emit: fastq
+    path  "*.version.txt"         , emit: version
+
+    script:
+    def software = getSoftwareName(task.process)
+    """
+    samtools \\
+        fastq \\
+        $options.args \\
+        -@ $task.cpus \\
+        $bam \\
+        > ${bam}.fastq
+
+    echo \$(samtools --version 2>&1) | sed 's/^.*samtools //; s/Using.*\$//' > ${software}.version.txt
+    """
+}

software/samtools/fastq/meta.yml

@@ -0,0 +1,43 @@
+name: samtools_fastq
+description: Converts a SAM/BAM/CRAM file to FASTA or FASTQ
+keywords:
+    - bam
+    - sam
+    - cram
+    - fasta
+    - fastq
+tools:
+    - samtools:
+        description: |
+            SAMtools is a set of utilities for interacting with and post-processing
+            short DNA sequence read alignments in the SAM, BAM and CRAM formats, written by Heng Li.
+            These files are generated as output by short read aligners like BWA.
+        homepage: http://www.htslib.org/
+        documentation: hhttp://www.htslib.org/doc/samtools.html
+        doi: 10.1093/bioinformatics/btp352
+input:
+    - meta:
+        type: map
+        description: |
+            Groovy Map containing sample information
+            e.g. [ id:'test', single_end:false ]
+    - bam:
+        type: file
+        description: BAM/CRAM/SAM file
+        pattern: "*.{bam,cram,sam}"
+output:
+    - meta:
+        type: map
+        description: |
+            Groovy Map containing sample information
+            e.g. [ id:'test', single_end:false ]
+    - fasta:
+        type: file
+        description: FASTA/FASTQ file
+        pattern: "*.{fasta,fastq}"
+    - version:
+        type: file
+        description: File containing software version
+        pattern: "*.{version.txt}"
+authors:
+  - "@suzannejin"

tests/config/pytest_software.yml

@@ -307,6 +307,10 @@ samtools_faidx:
   - software/samtools/faidx/**
   - tests/software/samtools/faidx/**
 
+samtools_fastq:
+  - software/samtools/fastq/**
+  - tests/software/samtools/fastq/**
+
 samtools_flagstat:
   - software/samtools/flagstat/**
   - tests/software/samtools/flagstat/**

tests/software/samtools/fastq/main.nf

@@ -0,0 +1,14 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { SAMTOOLS_FASTQ } from '../../../../software/samtools/fastq/main.nf' addParams( options: [:] )
+
+workflow test_samtools_fastq {
+    
+    def input = []
+    input = [ [ id:'test', single_end:false ], // meta map
+              file("${launchDir}/tests/data/genomics/sarscov2/bam/test_paired_end.sorted.bam", checkIfExists: true) ]
+
+    SAMTOOLS_FASTQ ( input )
+}

tests/software/samtools/fastq/test.yml

@@ -0,0 +1,8 @@
+- name: samtools fastq test_samtools_fastq
+  command: nextflow run tests/software/samtools/fastq -entry test_samtools_fastq -c tests/config/nextflow.config
+  tags:
+    - samtools_fastq
+    - samtools
+  files:
+    - path: output/samtools/test_paired_end.sorted.bam.fastq
+      md5sum: 4863ac55a2781962dba179a929673535