Merge pull request #89 from KevinMenden/master

Add bowtie module

.github/workflows/bowtie_align.yml

@@ -0,0 +1,42 @@
+name: bowtie_align
+on:
+    push:
+        paths:
+            - software/bowtie/align/**
+            - software/bowtie/index/**
+            - .github/workflows/bowtie_align.yml
+            - tests/software/bowtie/**
+    pull_request:
+        paths:
+            - software/bowtie/align/**
+            - software/bowtie/index/**
+            - .github/workflows/bowtie_align.yml
+            - tests/software/bowtie/**
+
+jobs:
+    ci_test:
+        runs-on: ubuntu-latest
+        strategy:
+            matrix:
+                nxf_version: [20.11.0-edge]
+        env:
+            NXF_ANSI_LOG: false
+        steps:
+            - uses: actions/checkout@v2
+
+            - name: Install Nextflow
+              env:
+                  NXF_VER: ${{ matrix.nxf_version }}
+              run: |
+                  wget -qO- get.nextflow.io | bash
+                  sudo mv nextflow /usr/local/bin/
+
+            - name: Set up Python
+              uses: actions/setup-python@v2
+              with:
+                  python-version: "3.x"
+            - name: Install dependencies
+              run: python -m pip install --upgrade pip pytest-workflow
+
+            # Test the module
+            - run: pytest --tag bowtie_align --symlink --wt 2

.github/workflows/bowtie_index.yml

@@ -0,0 +1,40 @@
+name: bowtie_index
+on:
+  push:
+    paths:
+      - software/bowtie/index/**
+      - .github/workflows/bowtie_index.yml
+      - tests/software/bowtie/**
+  pull_request:
+    paths:
+      - software/bowtie/index/**
+      - .github/workflows/bowtie_index.yml
+      - tests/software/bowtie/**
+
+jobs:
+  ci_test:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        nxf_version: [20.11.0-edge]
+    env:
+      NXF_ANSI_LOG: false
+    steps:
+      - uses: actions/checkout@v2
+
+      - name: Install Nextflow
+        env:
+          NXF_VER: ${{ matrix.nxf_version }}
+        run: |
+          wget -qO- get.nextflow.io | bash
+          sudo mv nextflow /usr/local/bin/
+
+      - name: Set up Python
+        uses: actions/setup-python@v2
+        with:
+          python-version: "3.x"
+      - name: Install dependencies
+        run: python -m pip install --upgrade pip pytest-workflow
+
+      # Test the module
+      - run: pytest --tag bowtie_index --symlink --wt 2

software/bowtie/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/bowtie/align/main.nf

@@ -0,0 +1,59 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName } from './functions'
+
+params.options = [:]
+def options    = initOptions(params.options)
+
+process BOWTIE_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::bowtie=1.3.0 bioconda::samtools=1.10' : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container 'https://depot.galaxyproject.org/singularity/mulled-v2-ffbf83a6b0ab6ec567a336cf349b80637135bca3:9e14e16c284d6860574cf5b624bbc44c793cb024-0'
+    } else {
+        container 'quay.io/biocontainers/mulled-v2-ffbf83a6b0ab6ec567a336cf349b80637135bca3:9e14e16c284d6860574cf5b624bbc44c793cb024-0'
+    }
+
+    input:
+    tuple val(meta), path(reads)
+    path  index
+
+    output:
+    tuple val(meta), path('*.bam'), emit: bam
+    tuple val(meta), path('*.out'), 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}"
+    def unaligned = params.save_unaligned ? "--un ${prefix}.unmapped.fastq" : ''
+    def endedness = meta.single_end ? "$reads" : "-1 ${reads[0]} -2 ${reads[1]}"
+    """
+    INDEX=`find -L ./ -name "*.3.ebwt" | sed 's/.3.ebwt//'`
+    bowtie \\
+        --threads $task.cpus \\
+        --sam \\
+        -x \$INDEX \\
+        -q \\
+        $unaligned \\
+        $options.args \\
+        $endedness \\
+        2> ${prefix}.out \\
+        | samtools view $options.args2 -@ $task.cpus -bS -o ${prefix}.bam -
+
+    if [ -f ${prefix}.unmapped.fastq ]; then
+        gzip ${prefix}.unmapped.fastq
+    fi
+    if [ -f ${prefix}.unmapped_1.fastq ]; then
+        gzip ${prefix}.unmapped_1.fastq
+        gzip ${prefix}.unmapped_2.fastq
+    fi
+
+    echo \$(bowtie --version 2>&1) | sed 's/^.*bowtie-align-s version //; s/ .*\$//' > ${software}.version.txt
+    """
+}

software/bowtie/align/meta.yml

@@ -0,0 +1,69 @@
+name: bowtie_align
+description: Align reads to a reference genome using bowtie
+keywords:
+    - align
+    - fasta
+    - genome
+    - reference
+tools:
+    - bowtie:
+          description: |
+              bowtie is a software package for mapping DNA sequences against
+              a large reference genome, such as the human genome.
+          homepage: http://bowtie-bio.sourceforge.net/index.shtml
+          documentation: http://bowtie-bio.sourceforge.net/manual.shtml
+          arxiv: arXiv:1303.3997
+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: Bowtie 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"
+authors:
+    - "@kevinmenden"

software/bowtie/index/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/bowtie/index/main.nf

@@ -0,0 +1,35 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName } from './functions'
+
+params.options = [:]
+def options    = initOptions(params.options)
+
+process BOWTIE_INDEX {
+    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::bowtie=1.3.0' : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container 'https://depot.galaxyproject.org/singularity/bowtie:1.3.0--py38hed8969a_1'
+    } else {
+        container 'quay.io/biocontainers/bowtie:1.3.0--py38hed8969a_1'
+    }
+
+    input:
+    path fasta
+
+    output:
+    path 'bowtie', emit: index
+    path '*.version.txt', emit: version
+
+    script:
+    def software  = getSoftwareName(task.process)
+    """
+    mkdir bowtie
+    bowtie-build --threads $task.cpus $fasta bowtie/${fasta.baseName}
+    echo \$(bowtie --version 2>&1) | sed 's/^.*bowtie-align-s version //; s/ .*\$//' > ${software}.version.txt
+    """
+}

software/bowtie/index/meta.yml

@@ -0,0 +1,51 @@
+name: bowtie_index
+description: Create bowtie index for reference genome
+keywords:
+    - index
+    - fasta
+    - genome
+    - reference
+tools:
+    - bowtie:
+          description: |
+              bowtie is a software package for mapping DNA sequences against
+              a large reference genome, such as the human genome.
+          homepage: http://bowtie-bio.sourceforge.net/index.shtml
+          documentation: http://bowtie-bio.sourceforge.net/manual.shtml
+          arxiv: arXiv:1303.3997
+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: Bowtie genome index files
+          pattern: "*.ebwt"
+    - version:
+          type: file
+          description: File containing software version
+          pattern: "*.{version.txt}"
+authors:
+    - "@kevinmenden"

tests/software/bowtie/main.nf

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

tests/software/bowtie/test.yml

@@ -0,0 +1,64 @@
+- name: Run bowtie index
+  command: nextflow run ./tests/software/bowtie -profile docker -entry test_bowtie_index -c tests/config/nextflow.config
+  tags:
+    - bowtie
+    - bowtie_index
+  files:
+    - path: output/bowtie/bowtie/NC_010473.1.ebwt
+      md5sum: 90f0b7aa5bbaeaaa999839ac13ad203c
+    - path: output/bowtie/bowtie/NC_010473.2.ebwt
+      md5sum: bfd10c5319c6a0dbc540fd789254a5dd
+    - path: output/bowtie/bowtie/NC_010473.3.ebwt
+      md5sum: cd201e81724f3099131aec16ef2cc53b
+    - path: output/bowtie/bowtie/NC_010473.4.ebwt
+      md5sum: bbb9d6d21ad765d135f95290204e8433
+    - path: output/bowtie/bowtie/NC_010473.rev.1.ebwt
+      md5sum: 44f719c2fe42e1f35d54e798775846d1
+    - path: output/bowtie/bowtie/NC_010473.rev.2.ebwt
+      md5sum: f3c398bba5158f4039334a932d79c051
+
+- name: Run bowtie index and align single-end
+  command: nextflow run ./tests/software/bowtie -profile docker -entry test_bowtie_alignment_single_end -c tests/config/nextflow.config
+  tags:
+    - bowtie
+    - bowtie_align
+  files:
+    - path: output/bowtie/bowtie/NC_010473.1.ebwt
+      md5sum: 90f0b7aa5bbaeaaa999839ac13ad203c
+    - path: output/bowtie/bowtie/NC_010473.2.ebwt
+      md5sum: bfd10c5319c6a0dbc540fd789254a5dd
+    - path: output/bowtie/bowtie/NC_010473.3.ebwt
+      md5sum: cd201e81724f3099131aec16ef2cc53b
+    - path: output/bowtie/bowtie/NC_010473.4.ebwt
+      md5sum: bbb9d6d21ad765d135f95290204e8433
+    - path: output/bowtie/bowtie/NC_010473.rev.1.ebwt
+      md5sum: 44f719c2fe42e1f35d54e798775846d1
+    - path: output/bowtie/bowtie/NC_010473.rev.2.ebwt
+      md5sum: f3c398bba5158f4039334a932d79c051
+    - path: output/bowtie/test.out
+      md5sum: a81cb18024616415a6cec3108a36fccd
+    - path: output/bowtie/test.bam
+      md5sum: 9feed8a55d4b5e600dcc577768ef07fc
+
+- name: Run bowtie index and align paired-end
+  command: nextflow run ./tests/software/bowtie -profile docker -entry test_bowtie_alignment_paired_end -c tests/config/nextflow.config
+  tags:
+    - bowtie
+    - bowtie_align
+  files:
+    - path: output/bowtie/bowtie/NC_010473.1.ebwt
+      md5sum: 90f0b7aa5bbaeaaa999839ac13ad203c
+    - path: output/bowtie/bowtie/NC_010473.2.ebwt
+      md5sum: bfd10c5319c6a0dbc540fd789254a5dd
+    - path: output/bowtie/bowtie/NC_010473.3.ebwt
+      md5sum: cd201e81724f3099131aec16ef2cc53b
+    - path: output/bowtie/bowtie/NC_010473.4.ebwt
+      md5sum: bbb9d6d21ad765d135f95290204e8433
+    - path: output/bowtie/bowtie/NC_010473.rev.1.ebwt
+      md5sum: 44f719c2fe42e1f35d54e798775846d1
+    - path: output/bowtie/bowtie/NC_010473.rev.2.ebwt
+      md5sum: f3c398bba5158f4039334a932d79c051
+    - path: output/bowtie/test.out
+      md5sum: a23e9a2a76e949aeb3693bcfae41a615
+    - path: output/bowtie/test.bam
+      md5sum: cf6a6381aa504e8342638ff3b509721e