Samtools fixmate module (#991)

* Samtools fixmate module

* Update modules/samtools/fixmate/meta.yml

Co-authored-by: James A. Fellows Yates <jfy133@gmail.com>

* Update modules/samtools/fixmate/main.nf

Co-authored-by: James A. Fellows Yates <jfy133@gmail.com>

* Update main.nf

* Update modules/samtools/fixmate/meta.yml

Co-authored-by: James A. Fellows Yates <jfy133@gmail.com>

* Update meta.yml

Co-authored-by: Simon Pearce <simon.pearce@cruk.manchester.ac.uk>
Co-authored-by: James A. Fellows Yates <jfy133@gmail.com>

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+    // Do not publish versions.yml unless running from pytest workflow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

@@ -0,0 +1,45 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName; getProcessName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process SAMTOOLS_FIXMATE {
+    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::samtools=1.14" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/samtools:1.14--hb421002_0"
+    } else {
+        container "quay.io/biocontainers/samtools:1.14--hb421002_0"
+    }
+
+    input:
+    tuple val(meta), path(bam)
+
+    output:
+    tuple val(meta), path("*.bam"), emit: bam
+    path "versions.yml"           , emit: versions
+
+    script:
+    def prefix = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    if ("$bam" == "${prefix}.bam") error "Input and output names are the same, use the suffix option to disambiguate!"
+
+    """
+    samtools \\
+        fixmate  \\
+        $options.args \\
+        -@ $task.cpus \\
+        $bam \\
+        ${prefix}.bam \\
+
+    cat <<-END_VERSIONS > versions.yml
+    ${getProcessName(task.process)}:
+        ${getSoftwareName(task.process)}: \$(echo \$(samtools --version 2>&1) | sed 's/^.*samtools //; s/Using.*\$//')
+    END_VERSIONS
+    """
+}

modules/samtools/fixmate/meta.yml

@@ -0,0 +1,49 @@
+name: samtools_fixmate
+description: Samtools fixmate is a tool that can fill in information (insert size, cigar, mapq) about paired end reads onto the corresponding other read. Also has options to remove secondary/unmapped alignments and recalculate whether reads are proper pairs.
+keywords:
+    - fixmate
+    - samtools
+    - insert size
+    - repair
+    - bam
+    - paired
+    - read pairs
+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: http://www.htslib.org/doc/samtools.html
+        tool_dev_url: https://github.com/samtools/samtools
+        doi: 10.1093/bioinformatics/btp352
+        licence: ['MIT']
+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, must be sorted by name, not coordinate
+      pattern: "*.{bam,cram,sam}"
+
+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"
+  - bam:
+      type: file
+      description: A BAM/CRAM/SAM file with mate information added and/or proper pairs recalled
+      pattern: "*.{bam,cram,sam}"
+
+authors:
+  - "@sppearce"

tests/config/pytest_modules.yml

@@ -1027,6 +1027,10 @@ samtools/fastq:
   - modules/samtools/fastq/**
   - tests/modules/samtools/fastq/**
 
+samtools/fixmate:
+  - modules/samtools/fixmate/**
+  - tests/modules/samtools/fixmate/**
+
 samtools/flagstat:
   - modules/samtools/flagstat/**
   - tests/modules/samtools/flagstat/**

tests/modules/samtools/fixmate/main.nf

@@ -0,0 +1,14 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { SAMTOOLS_FIXMATE } from '../../../../modules/samtools/fixmate/main.nf' addParams( options: [args:'-r -c -m'] )
+
+workflow test_samtools_fixmate {
+
+    input = [ [ id:'test', single_end:false ], // meta map
+              file(params.test_data['sarscov2']['illumina']['test_paired_end_bam'], checkIfExists: true) ]
+
+    SAMTOOLS_FIXMATE ( input )
+
+}

tests/modules/samtools/fixmate/test.yml

@@ -0,0 +1,8 @@
+- name: samtools fixmate test_samtools_fixmate
+  command: nextflow run tests/modules/samtools/fixmate -entry test_samtools_fixmate -c tests/config/nextflow.config
+  tags:
+    - samtools/fixmate
+    - samtools
+  files:
+    - path: output/samtools/test.bam
+      md5sum: 92c8463710cdcaef2010aa02ed9e01fd