Add Medaka module (#992)

* add racon

* add medaka module

* add medaka module

* add medaka module

* add medaka module

* add medaka module

* add medaka module

* Indentation

* Apply suggestions from code review

Co-authored-by: FriederikeHanssen <Friederike.hanssen@qbic.uni-tuebingen.de>

* Update main.nf

* Update main.nf

* Apply suggestions from code review

Co-authored-by: FriederikeHanssen <Friederike.hanssen@qbic.uni-tuebingen.de>
Co-authored-by: Harshil Patel <drpatelh@users.noreply.github.com>

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+    // Do not publish versions.yml unless running from pytest workflow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

@@ -0,0 +1,47 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName; getProcessName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process MEDAKA {
+    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::medaka=1.4.4" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/medaka:1.4.4--py38h130def0_0"
+    } else {
+        container "quay.io/biocontainers/medaka:1.4.4--py38h130def0_0"
+    }
+
+    input:
+    tuple val(meta), path(reads), path(assembly)
+
+    output:
+    tuple val(meta), path("*.fa.gz"), emit: assembly
+    path "versions.yml"             , emit: versions
+
+    script:
+    def prefix = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    """
+    medaka_consensus \\
+        -t $task.cpus \\
+        $options.args \\
+        -i $reads \\
+        -d $assembly \\
+        -o ./
+
+    mv consensus.fasta ${prefix}.fa
+
+    gzip -n ${prefix}.fa
+
+    cat <<-END_VERSIONS > versions.yml
+    ${getProcessName(task.process)}:
+        ${getSoftwareName(task.process)}: \$( medaka --version 2>&1 | sed 's/medaka //g' )
+    END_VERSIONS
+    """
+}

modules/medaka/meta.yml

@@ -0,0 +1,47 @@
+name: medaka
+description: A tool to create consensus sequences and variant calls from nanopore sequencing data
+keywords:
+  - assembly
+  - polishing
+  - nanopore
+tools:
+  - medaka:
+      description: Neural network sequence error correction.
+      homepage: https://nanoporetech.github.io/medaka/index.html
+      documentation: https://nanoporetech.github.io/medaka/index.html
+      tool_dev_url: https://github.com/nanoporetech/medaka
+      doi: ""
+      licence: ['Mozilla Public License 2.0']
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - reads:
+      type: file
+      description: List of input nanopore fasta/FastQ files
+      pattern: "*.{fasta,fa,fastq,fastq.gz,fq,fq.gz}"
+  - assembly:
+      type: file
+      description: Genome assembly
+      pattern: "*.{fasta,fa}"
+
+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"
+  - assembly:
+      type: file
+      description: Polished genome assembly
+      pattern: "*.fa.gz"
+
+authors:
+  - "@avantonder"

tests/config/pytest_modules.yml

@@ -822,6 +822,10 @@ maxbin2:
   - modules/maxbin2/**
   - tests/modules/maxbin2/**
 
+medaka:
+  - modules/medaka/**
+  - tests/modules/medaka/**
+
 megahit:
   - modules/megahit/**
   - tests/modules/megahit/**

tests/modules/medaka/main.nf

@@ -0,0 +1,16 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { MEDAKA } from '../../../modules/medaka/main.nf' addParams( options: [suffix:'.polished.genome'] )
+
+workflow test_medaka {
+    
+    input = [ 
+        [ id:'test', single_end:true ], // meta map
+        file(params.test_data['sarscov2']['nanopore']['test_fastq_gz'], checkIfExists: true),
+        file(params.test_data['sarscov2']['genome']['genome_fasta'], checkIfExists: true)
+    ]
+    
+    MEDAKA ( input )
+}

tests/modules/medaka/test.yml

@@ -0,0 +1,7 @@
+- name: medaka test_medaka
+  command: nextflow run ./tests/modules/medaka -entry test_medaka -c tests/config/nextflow.config
+  tags:
+    - medaka
+  files:
+    - path: output/medaka/test.polished.genome.fa.gz
+      md5sum: f42303f1d6c2c79175faeb00e10b9a6e