add filtlong (#919)

modules/filtlong/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+    // Do not publish versions.yml unless running from pytest workflow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/filtlong/main.nf

@@ -0,0 +1,43 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName; getProcessName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process FILTLONG {
+    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::filtlong=0.2.1" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/filtlong:0.2.1--h9a82719_0"
+    } else {
+        container "quay.io/biocontainers/filtlong:0.2.1--h9a82719_0"
+    }
+
+    input:
+    tuple val(meta), path(shortreads), path(longreads)
+
+    output:
+    tuple val(meta), path("${meta.id}_lr_filtlong.fastq.gz"), emit: reads
+    path "versions.yml"                                     , emit: versions
+
+    script:
+    def prefix = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    def short_reads = meta.single_end ? "-1 $shortreads" : "-1 ${shortreads[0]} -2 ${shortreads[1]}"
+    """
+    filtlong \\
+        $short_reads \\
+        $options.args \\
+        $longreads \\
+        | gzip -n > ${prefix}_lr_filtlong.fastq.gz
+
+    cat <<-END_VERSIONS > versions.yml
+    ${getProcessName(task.process)}:
+        ${getSoftwareName(task.process)}: \$( filtlong --version | sed -e "s/Filtlong v//g" )
+    END_VERSIONS
+    """
+}

modules/filtlong/meta.yml

@@ -0,0 +1,50 @@
+name: filtlong
+description: Filtlong filters long reads based on quality measures or short read data.
+keywords:
+  - nanopore
+  - quality control
+  - QC
+  - filtering
+  - long reads
+  - short reads
+tools:
+  - filtlong:
+      description: Filtlong is a tool for filtering long reads. It can take a set of long reads and produce a smaller, better subset. It uses both read length (longer is better) and read identity (higher is better) when choosing which reads pass the filter.
+      homepage: https://anaconda.org/bioconda/filtlong
+      documentation: None
+      tool_dev_url: https://github.com/rrwick/Filtlong
+      doi: ""
+      licence: ['GPL v3']
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - shortreads:
+      type: file
+      description: fastq file
+      pattern: "*.{fq,fastq,fq.gz,fastq.gz}"
+  - longreads:
+      type: file
+      description: fastq file
+      pattern: "*.{fq,fastq,fq.gz,fastq.gz}"
+
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+  - reads:
+      type: file
+      description: Filtered (compressed) fastq file
+      pattern: "*.fastq.gz"
+
+authors:
+  - "@d4straub"

tests/config/pytest_modules.yml

@@ -382,6 +382,10 @@ fgbio/sortbam:
   - modules/fgbio/sortbam/**
   - tests/modules/fgbio/sortbam/**
 
+filtlong:
+  - modules/filtlong/**
+  - tests/modules/filtlong/**
+
 flash:
   - modules/flash/**
   - tests/modules/flash/**

tests/modules/filtlong/main.nf

@@ -0,0 +1,36 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { FILTLONG } from '../../../modules/filtlong/main.nf' addParams( options: [:] )
+
+workflow test_filtlong {
+
+    input = [ [ id:'test', single_end:false ], // meta map
+              [],
+              [ file(params.test_data['sarscov2']['nanopore']['test_fastq_gz'], checkIfExists: true) ]
+            ]
+
+    FILTLONG ( input )
+}
+
+workflow test_filtlong_illumina_se {
+
+    input = [ [ id:'test', single_end:true ], // meta map
+              [ file(params.test_data['sarscov2']['illumina']['test_1_fastq_gz'], checkIfExists: true) ],
+              [ file(params.test_data['sarscov2']['nanopore']['test_fastq_gz'], checkIfExists: true) ]
+            ]
+
+    FILTLONG ( input )
+}
+
+workflow test_filtlong_illumina_pe {
+
+    input = [ [ id:'test', single_end:false ], // meta map
+              [ file(params.test_data['sarscov2']['illumina']['test_1_fastq_gz'], checkIfExists: true),
+                file(params.test_data['sarscov2']['illumina']['test_2_fastq_gz'], checkIfExists: true) ],
+              [ file(params.test_data['sarscov2']['nanopore']['test_fastq_gz'], checkIfExists: true) ]
+            ]
+
+    FILTLONG ( input )
+}

tests/modules/filtlong/test.yml

@@ -0,0 +1,23 @@
+- name: filtlong test_filtlong
+  command: nextflow run tests/modules/filtlong -entry test_filtlong -c tests/config/nextflow.config
+  tags:
+    - filtlong
+  files:
+    - path: output/filtlong/test_lr_filtlong.fastq.gz
+      md5sum: 7029066c27ac6f5ef18d660d5741979a
+
+- name: filtlong test_filtlong_illumina_se
+  command: nextflow run tests/modules/filtlong -entry test_filtlong_illumina_se -c tests/config/nextflow.config
+  tags:
+    - filtlong
+  files:
+    - path: output/filtlong/test_lr_filtlong.fastq.gz
+      md5sum: 7029066c27ac6f5ef18d660d5741979a
+
+- name: filtlong test_filtlong_illumina_pe
+  command: nextflow run tests/modules/filtlong -entry test_filtlong_illumina_pe -c tests/config/nextflow.config
+  tags:
+    - filtlong
+  files:
+    - path: output/filtlong/test_lr_filtlong.fastq.gz
+      md5sum: 7029066c27ac6f5ef18d660d5741979a