Merge pull request #127 from JoseEspinosa/bowtie2_viralrecon

Bowtie2 viralrecon

software/bowtie2/align/functions.nf

@@ -0,0 +1,60 @@
+
+/*
+ * -----------------------------------------------------
+ *  Utility functions used in nf-core DSL2 module files
+ * -----------------------------------------------------
+ */
+
+/*
+ * Extract name of software tool from process name using $task.process
+ */
+def getSoftwareName(task_process) {
+    return task_process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()
+}
+
+/*
+ * Function to initialise default values and to generate a Groovy Map of available options for nf-core modules
+ */
+def initOptions(Map args) {
+    def Map options = [:]
+    options.args          = args.args ?: ''
+    options.args2         = args.args2 ?: ''
+    options.publish_by_id = args.publish_by_id ?: false
+    options.publish_dir   = args.publish_dir ?: ''
+    options.publish_files = args.publish_files
+    options.suffix        = args.suffix ?: ''
+    return options
+}
+
+/*
+ * Tidy up and join elements of a list to return a path string
+ */
+def getPathFromList(path_list) {
+    def paths = path_list.findAll { item -> !item?.trim().isEmpty() }  // Remove empty entries
+    paths = paths.collect { it.trim().replaceAll("^[/]+|[/]+\$", '') } // Trim whitespace and trailing slashes
+    return paths.join('/')
+}
+
+/*
+ * Function to save/publish module results
+ */
+def saveFiles(Map args) {
+    if (!args.filename.endsWith('.version.txt')) {
+        def ioptions = initOptions(args.options)
+        def path_list = [ ioptions.publish_dir ?: args.publish_dir ]
+        if (ioptions.publish_by_id) {
+            path_list.add(args.publish_id)
+        }
+        if (ioptions.publish_files instanceof Map) {
+            for (ext in ioptions.publish_files) {
+                if (args.filename.endsWith(ext.key)) {
+                    def ext_list = path_list.collect()
+                    ext_list.add(ext.value)
+                    return "${getPathFromList(ext_list)}/$args.filename"
+                }
+            }
+        } else if (ioptions.publish_files == null) {
+            return "${getPathFromList(path_list)}/$args.filename"
+        }
+    }
+}

software/bowtie2/align/main.nf

@@ -0,0 +1,72 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName } from './functions'
+
+params.options = [:]
+def options    = initOptions(params.options)
+
+process BOWTIE2_ALIGN {
+    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), publish_id:meta.id) }
+
+    conda (params.enable_conda ? 'bioconda::bowtie2=2.4.2 bioconda::samtools=1.11 conda-forge::pigz=2.3.4' : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/mulled-v2-ac74a7f02cebcfcc07d8e8d1d750af9c83b4d45a:577a697be67b5ae9b16f637fd723b8263a3898b3-0"
+    } else {
+        container "quay.io/biocontainers/mulled-v2-ac74a7f02cebcfcc07d8e8d1d750af9c83b4d45a:577a697be67b5ae9b16f637fd723b8263a3898b3-0"
+    }
+
+    input:
+    tuple val(meta), path(reads)
+    path  index
+
+    output:
+    tuple val(meta), path('*.bam'), emit: bam
+    tuple val(meta), path('*.log'), emit: log
+    path  '*.version.txt'         , emit: version
+    tuple val(meta), path('*fastq.gz'), optional:true, emit: fastq
+
+    script:
+    def software = getSoftwareName(task.process)
+    def prefix   = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    if (meta.single_end) {
+        def unaligned = params.save_unaligned ? "--un-gz ${prefix}.unmapped.fastq.gz" : ''
+        """
+        INDEX=`find -L ./ -name "*.rev.1.bt2" | sed 's/.rev.1.bt2//'`
+        bowtie2 \\
+            -x \$INDEX \\
+            -U $reads \\
+            --threads $task.cpus \\
+            $unaligned \\
+            $options.args \\
+            2> ${prefix}.bowtie2.log \\
+            | samtools view -@ $task.cpus $options.args2 -bhS -o ${prefix}.bam -
+
+        echo \$(bowtie2 --version 2>&1) | sed 's/^.*bowtie2-align-s version //; s/ .*\$//' > ${software}.version.txt
+        """
+    } else {
+        def unaligned = params.save_unaligned ? "--un-conc-gz ${prefix}.unmapped.fastq.gz" : ''
+        """
+        INDEX=`find -L ./ -name "*.rev.1.bt2" | sed 's/.rev.1.bt2//'`
+        bowtie2 \\
+            -x \$INDEX \\
+            -1 ${reads[0]} \\
+            -2 ${reads[1]} \\
+            --threads $task.cpus \\
+            $unaligned \\
+            $options.args \\
+            2> ${prefix}.bowtie2.log \\
+            | samtools view -@ $task.cpus $options.args2 -bhS -o ${prefix}.bam -
+
+        if [ -f ${prefix}.unmapped.fastq.1.gz ]; then
+            mv ${prefix}.unmapped.fastq.1.gz ${prefix}.unmapped_1.fastq.gz
+        fi
+        if [ -f ${prefix}.unmapped.fastq.2.gz ]; then
+            mv ${prefix}.unmapped.fastq.2.gz ${prefix}.unmapped_2.fastq.gz
+        fi
+        echo \$(bowtie2 --version 2>&1) | sed 's/^.*bowtie2-align-s version //; s/ .*\$//' > ${software}.version.txt
+        """
+    }
+}

software/bowtie2/align/meta.yml

@@ -0,0 +1,74 @@
+name: bowtie2_align
+description: Align reads to a reference genome using bowtie2
+keywords:
+    - align
+    - fasta
+    - genome
+    - reference
+tools:
+    - bowtie2:
+          description: |
+              Bowtie 2 is an ultrafast and memory-efficient tool for aligning
+              sequencing reads to long reference sequences.
+          homepage: http://bowtie-bio.sourceforge.net/bowtie2/index.shtml
+          documentation: http://bowtie-bio.sourceforge.net/bowtie2/manual.shtml
+          doi: 10.1038/nmeth.1923
+params:
+    - outdir:
+          type: string
+          description: |
+              The pipeline's output directory. By default, the module will
+              output files into `$params.outdir/<SOFTWARE>`
+    - publish_dir_mode:
+          type: string
+          description: |
+              Value for the Nextflow `publishDir` mode parameter.
+              Available: symlink, rellink, link, copy, copyNoFollow, move.
+    - enable_conda:
+          type: boolean
+          description: |
+              Run the module with Conda using the software specified
+              via the `conda` directive
+    - singularity_pull_docker_container:
+          type: boolean
+          description: |
+              Instead of directly downloading Singularity images for use with Singularity,
+              force the workflow to pull and convert Docker containers instead.
+    - save_unaligned:
+          type: boolean
+          description: Save unaligned reads
+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 1 and 2 for single-end and paired-end data,
+              respectively.
+    - index:
+          type: file
+          description: Bowtie2 genome index files
+          pattern: "*.ebwt"
+output:
+    - bam:
+          type: file
+          description: Output BAM file containing read alignments
+          pattern: "*.{bam}"
+    - version:
+          type: file
+          description: File containing software version
+          pattern: "*.{version.txt}"
+    - fastq:
+          type: file
+          description: Unaligned FastQ files
+          pattern: "*.fastq.gz"
+    - log:
+        type: file
+        description: Aligment log
+        pattern: "*.log"
+authors:
+    - "@joseespinosa"
+    - "@drpatelh"

software/bowtie2/build/functions.nf

@@ -0,0 +1,60 @@
+
+/*
+ * -----------------------------------------------------
+ *  Utility functions used in nf-core DSL2 module files
+ * -----------------------------------------------------
+ */
+
+/*
+ * Extract name of software tool from process name using $task.process
+ */
+def getSoftwareName(task_process) {
+    return task_process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()
+}
+
+/*
+ * Function to initialise default values and to generate a Groovy Map of available options for nf-core modules
+ */
+def initOptions(Map args) {
+    def Map options = [:]
+    options.args          = args.args ?: ''
+    options.args2         = args.args2 ?: ''
+    options.publish_by_id = args.publish_by_id ?: false
+    options.publish_dir   = args.publish_dir ?: ''
+    options.publish_files = args.publish_files
+    options.suffix        = args.suffix ?: ''
+    return options
+}
+
+/*
+ * Tidy up and join elements of a list to return a path string
+ */
+def getPathFromList(path_list) {
+    def paths = path_list.findAll { item -> !item?.trim().isEmpty() }  // Remove empty entries
+    paths = paths.collect { it.trim().replaceAll("^[/]+|[/]+\$", '') } // Trim whitespace and trailing slashes
+    return paths.join('/')
+}
+
+/*
+ * Function to save/publish module results
+ */
+def saveFiles(Map args) {
+    if (!args.filename.endsWith('.version.txt')) {
+        def ioptions = initOptions(args.options)
+        def path_list = [ ioptions.publish_dir ?: args.publish_dir ]
+        if (ioptions.publish_by_id) {
+            path_list.add(args.publish_id)
+        }
+        if (ioptions.publish_files instanceof Map) {
+            for (ext in ioptions.publish_files) {
+                if (args.filename.endsWith(ext.key)) {
+                    def ext_list = path_list.collect()
+                    ext_list.add(ext.value)
+                    return "${getPathFromList(ext_list)}/$args.filename"
+                }
+            }
+        } else if (ioptions.publish_files == null) {
+            return "${getPathFromList(path_list)}/$args.filename"
+        }
+    }
+}

software/bowtie2/build/main.nf

@@ -0,0 +1,35 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName } from './functions'
+
+params.options = [:]
+def options    = initOptions(params.options)
+
+process BOWTIE2_BUILD {
+    tag "$fasta"
+    label 'process_high'
+    publishDir "${params.outdir}",
+        mode: params.publish_dir_mode,
+        saveAs: { filename -> saveFiles(filename:filename, options:params.options, publish_dir:getSoftwareName(task.process), publish_id:'') }
+
+    conda (params.enable_conda ? 'bioconda::bowtie2=2.4.2' : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container 'https://depot.galaxyproject.org/singularity/bowtie2:2.4.2--py38h1c8e9b9_1'
+    } else {
+        container 'quay.io/biocontainers/bowtie2:2.4.2--py38h1c8e9b9_1'
+    }
+
+    input:
+    path fasta
+
+    output:
+    path 'bowtie2'      , emit: index
+    path '*.version.txt', emit: version
+
+    script:
+    def software  = getSoftwareName(task.process)
+    """
+    mkdir bowtie2
+    bowtie2-build $options.args --threads $task.cpus $fasta bowtie2/${fasta.baseName}
+    echo \$(bowtie2 --version 2>&1) | sed 's/^.*bowtie2-align-s version //; s/ .*\$//' > ${software}.version.txt
+    """
+}

software/bowtie2/build/meta.yml

@@ -0,0 +1,53 @@
+name: bowtie2_build
+description: Builds bowtie index for reference genome
+keywords:
+    - build
+    - index
+    - fasta
+    - genome
+    - reference
+tools:
+    - bowtie2:
+          description: |
+              Bowtie 2 is an ultrafast and memory-efficient tool for aligning
+              sequencing reads to long reference sequences.
+          homepage: http://bowtie-bio.sourceforge.net/bowtie2/index.shtml
+          documentation: http://bowtie-bio.sourceforge.net/bowtie2/manual.shtml
+          doi: 10.1038/nmeth.1923
+params:
+    - outdir:
+          type: string
+          description: |
+              The pipeline's output directory. By default, the module will
+              output files into `$params.outdir/<SOFTWARE>`
+    - publish_dir_mode:
+          type: string
+          description: |
+              Value for the Nextflow `publishDir` mode parameter.
+              Available: symlink, rellink, link, copy, copyNoFollow, move.
+    - enable_conda:
+          type: boolean
+          description: |
+              Run the module with Conda using the software specified
+              via the `conda` directive
+    - singularity_pull_docker_container:
+          type: boolean
+          description: |
+              Instead of directly downloading Singularity images for use with Singularity,
+              force the workflow to pull and convert Docker containers instead.
+input:
+    - fasta:
+          type: file
+          description: Input genome fasta file
+output:
+    - index:
+          type: file
+          description: Bowtie2 genome index files
+          pattern: "*.bt2"
+    - version:
+          type: file
+          description: File containing software version
+          pattern: "*.{version.txt}"
+authors:
+    - "@joseespinosa"
+    - "@drpatelh"

tests/software/bowtie2/main.nf

@@ -0,0 +1,33 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { BOWTIE2_BUILD } from '../../../software/bowtie2/build/main.nf' addParams( options: [:] )
+include { BOWTIE2_ALIGN } from '../../../software/bowtie2/align/main.nf'  addParams( options: [:] )
+
+workflow test_bowtie2_build {
+    fasta = file("${launchDir}/tests/data/fasta/E_coli/NC_010473.fa", checkIfExists: true)
+    BOWTIE2_BUILD ( fasta )
+}
+
+workflow test_bowtie2_alignment_single_end {
+
+    fasta = file("${launchDir}/tests/data/fasta/E_coli/NC_010473.fa", checkIfExists: true)
+    BOWTIE2_BUILD ( fasta )
+
+    def input = []
+    input = [ [ id:'test', single_end:true ], // meta map
+              [ file("${launchDir}/tests/data/fastq/rna/test_R1.fastq.gz", checkIfExists: true) ] ]
+    BOWTIE2_ALIGN ( input, BOWTIE2_BUILD.out.index )
+}
+
+workflow test_bowtie2_alignment_paired_end {
+    fasta = file("${launchDir}/tests/data/fasta/E_coli/NC_010473.fa", checkIfExists: true)
+    BOWTIE2_BUILD ( fasta )
+
+    def input = []
+    input = [ [ id:'test', single_end:false ], // meta map
+              [ file("${launchDir}/tests/data/fastq/rna/test_R1.fastq.gz", checkIfExists: true),
+                file("${launchDir}/tests/data/fastq/rna/test_R2.fastq.gz", checkIfExists: true) ] ]
+    BOWTIE2_ALIGN ( input, BOWTIE2_BUILD.out.index )
+}

tests/software/bowtie2/test.yml

@@ -0,0 +1,64 @@
+- name: Run bowtie2 build
+  command: nextflow run ./tests/software/bowtie2 -profile docker -entry test_bowtie2_build -c tests/config/nextflow.config
+  tags:
+    - bowtie2
+    - bowtie2_build
+  files:
+    - path: output/bowtie2/bowtie2/NC_010473.1.bt2
+      md5sum: 4db22d92e72111a5fbf609b3d9a43015
+    - path: output/bowtie2/bowtie2/NC_010473.2.bt2
+      md5sum: f4429ec74ee0064732c4bb26718a3eb1
+    - path: output/bowtie2/bowtie2/NC_010473.3.bt2
+      md5sum: cd201e81724f3099131aec16ef2cc53b
+    - path: output/bowtie2/bowtie2/NC_010473.4.bt2
+      md5sum: bbb9d6d21ad765d135f95290204e8433
+    - path: output/bowtie2/bowtie2/NC_010473.rev.1.bt2
+      md5sum: 4ccfee8857c3b1c69857e5ecdef597aa
+    - path: output/bowtie2/bowtie2/NC_010473.rev.2.bt2
+      md5sum: 5e8fb4af677eb3a57a40e76dc3f6db72
+
+- name: Run bowtie2 index and align single-end
+  command: nextflow run ./tests/software/bowtie2 -profile docker -entry test_bowtie2_alignment_single_end -c tests/config/nextflow.config
+  tags:
+    - bowtie2
+    - bowtie2_align
+  files:
+    - path: output/bowtie2/bowtie2/NC_010473.1.bt2
+      md5sum: 4db22d92e72111a5fbf609b3d9a43015
+    - path: output/bowtie2/bowtie2/NC_010473.2.bt2
+      md5sum: f4429ec74ee0064732c4bb26718a3eb1
+    - path: output/bowtie2/bowtie2/NC_010473.3.bt2
+      md5sum: cd201e81724f3099131aec16ef2cc53b
+    - path: output/bowtie2/bowtie2/NC_010473.4.bt2
+      md5sum: bbb9d6d21ad765d135f95290204e8433
+    - path: output/bowtie2/bowtie2/NC_010473.rev.1.bt2
+      md5sum: 4ccfee8857c3b1c69857e5ecdef597aa
+    - path: output/bowtie2/bowtie2/NC_010473.rev.2.bt2
+      md5sum: 5e8fb4af677eb3a57a40e76dc3f6db72
+    - path: output/bowtie2/test.bowtie2.log
+      md5sum: 90041c264231be535042adb93a279356
+    - path: output/bowtie2/test.bam
+      md5sum: 906102f401d8234b6473790988fb09cf
+
+- name: Run bowtie2 index and align paired-end
+  command: nextflow run ./tests/software/bowtie2 -profile docker -entry test_bowtie2_alignment_paired_end -c tests/config/nextflow.config
+  tags:
+    - bowtie2
+    - bowtie2_align
+  files:
+    - path: output/bowtie2/bowtie2/NC_010473.1.bt2
+      md5sum: 90f0b7aa5bbaeaaa999839ac13ad203c
+    - path: output/bowtie2/bowtie2/NC_010473.2.bt2
+      md5sum: bfd10c5319c6a0dbc540fd789254a5dd
+    - path: output/bowtie2/bowtie2/NC_010473.3.bt2
+      md5sum: cd201e81724f3099131aec16ef2cc53b
+    - path: output/bowtie2/bowtie2/NC_010473.4.bt2
+      md5sum: bbb9d6d21ad765d135f95290204e8433
+    - path: output/bowtie2/bowtie2/NC_010473.rev.1.bt2
+      md5sum: 44f719c2fe42e1f35d54e798775846d1
+    - path: output/bowtie2/bowtie2/NC_010473.rev.2.bt2
+      md5sum: f3c398bba5158f4039334a932d79c051
+    - path: output/bowtie2/test.bowtie2.log
+      md5sum: 9f9eb40b5f57e0f2d5c874f2c1b5cfd5
+    - path: output/bowtie2/test.bam
+      md5sum: 732a33cd9816b38ea9b919c25eeb78f8