add flash module (#341)

* add flash module

* remove todo

* run tests

Co-authored-by: Harshil Patel <drpatelh@users.noreply.github.com>

software/flash/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/flash/main.nf

@@ -0,0 +1,40 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process FLASH {
+    tag "$meta.id"
+    label 'process_medium'
+    publishDir "${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::flash=1.2.11" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/flash:1.2.11--hed695b0_5"
+    } else {
+        container "quay.io/biocontainers/flash:1.2.11--hed695b0_5"
+    }
+
+    input:
+    tuple val(meta), path(reads)
+
+    output:
+    tuple val(meta), path("*.merged.*.fastq.gz"), emit: reads
+    path "*.version.txt"                      , emit: version
+
+    script:
+    def software = getSoftwareName(task.process)
+    def prefix   = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    def merged   = "-o ${prefix}.merged"
+    def input_reads = "${reads[0]} ${reads[1]}"
+    """
+    flash \\
+        $options.args \\
+        $merged \\
+        -z \\
+        $input_reads
+    echo \$(flash --version) > ${software}.version.txt
+    """
+}

software/flash/meta.yml

@@ -0,0 +1,44 @@
+name: flash
+description: Perform merging of mate paired-end sequencing reads
+keywords:
+  - sort
+  - reads merging
+  - merge mate pairs
+tools:
+  - flash:
+      description: | 
+        Merge mates from fragments that are shorter than twice the read length
+      homepage: https://ccb.jhu.edu/software/FLASH/
+      documentation: {}
+      doi: 10.1093/bioinformatics/btr507
+      licence: ['GPL v3+']
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - reads:
+      type: file
+      description: |
+        List of input FastQ files of size 2; i.e., paired-end data.
+      pattern: "*fastq.gz"
+
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - reads:
+      type: file
+      description: The merged fastq reads
+      pattern: "*fastq.gz"
+  - version:
+      type: file
+      description: File containing software version
+      pattern: "*.{version.txt}"
+
+authors:
+  - "@Erkison"

tests/config/pytest_software.yml

@@ -178,6 +178,10 @@ fastqc:
   - software/fastqc/**
   - tests/software/fastqc/**
 
+flash:
+  - software/flash/**
+  - tests/software/flash/**
+
 gatk4_applybqsr:
   - software/gatk4/applybqsr/**
   - tests/software/gatk4/applybqsr/**

tests/software/flash/main.nf

@@ -0,0 +1,15 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { FLASH } from '../../../software/flash/main.nf' addParams( options: [args:'-m 20 -M 100'] )
+
+workflow test_flash {
+    
+    def input = []
+    input = [ [ id:'test', single_end:false ], // meta map
+              [ file("${launchDir}/tests/data/genomics/sarscov2/fastq/test_1.fastq.gz", checkIfExists: true),
+                file("${launchDir}/tests/data/genomics/sarscov2/fastq/test_2.fastq.gz", checkIfExists: true) ] ]
+
+    FLASH ( input )
+}

tests/software/flash/test.yml

@@ -0,0 +1,11 @@
+- name: flash test_flash
+  command: nextflow run tests/software/flash -entry test_flash -c tests/config/nextflow.config
+  tags:
+    - flash
+  files:
+    - path: output/flash/test.merged.notCombined_2.fastq.gz
+      md5sum: 96ec044281fe60e0061976d928810314
+    - path: output/flash/test.merged.extendedFrags.fastq.gz
+      md5sum: da20afa705e8ea881e66960bb75607c9
+    - path: output/flash/test.merged.notCombined_1.fastq.gz
+      md5sum: 32451c87f89172c764bec19136592d29