Merge pull request #126 from JoseEspinosa/samtools_viralrecon

Samtools viralrecon

.github/workflows/samtools_mpileup.yml

@@ -0,0 +1,40 @@
+name: samtools_mpileup
+on:
+  push:
+    paths:
+      - software/samtools/mpileup/**
+      - .github/workflows/samtools_mpileup.yml
+      - tests/software/samtools/**
+  pull_request:
+    paths:
+      - software/samtools/mpileup/**
+      - .github/workflows/samtools_mpileup.yml
+      - tests/software/samtools/**
+
+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 samtools_mpileup --symlink --wt 2

.github/workflows/samtools_view.yml

@@ -0,0 +1,40 @@
+name: samtools_view
+on:
+  push:
+    paths:
+      - software/samtools/view/**
+      - .github/workflows/samtools_view.yml
+      - tests/software/samtools/**
+  pull_request:
+    paths:
+      - software/samtools/view/**
+      - .github/workflows/samtools_view.yml
+      - tests/software/samtools/**
+
+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 samtools_view --symlink --wt 2

software/samtools/mpileup/functions.nf

@@ -0,0 +1,59 @@
+/*
+ * -----------------------------------------------------
+ *  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/samtools/mpileup/main.nf

@@ -0,0 +1,40 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName } from './functions'
+
+params.options = [:]
+def options    = initOptions(params.options)
+
+process SAMTOOLS_MPILEUP {
+    tag "$meta.id"
+    label 'process_medium'
+    publishDir "${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::samtools=1.10" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/samtools:1.10--h9402c20_2"
+    } else {
+        container "quay.io/biocontainers/samtools:1.10--h9402c20_2"
+    }
+
+    input:
+    tuple val(meta), path(bam)
+    path  fasta
+
+    output:
+    tuple val(meta), path("*.mpileup"), emit: mpileup
+    path  "*.version.txt"             , emit: version
+
+    script:
+    def software = getSoftwareName(task.process)
+    def prefix   = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    """
+    samtools mpileup \\
+        --fasta-ref $fasta \\
+        --output ${prefix}.mpileup \\
+        $options.args \\
+        $bam
+    echo \$(samtools --version 2>&1) | sed 's/^.*samtools //; s/Using.*\$//' > ${software}.version.txt
+    """
+}

software/samtools/mpileup/meta.yml

@@ -0,0 +1,68 @@
+name: samtools_mpileup
+description:  BAM
+keywords:
+    - mpileup
+    - bam
+    - sam
+    - cram
+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: hhttp://www.htslib.org/doc/samtools.html
+        doi: 10.1093/bioinformatics/btp352
+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:
+    - meta:
+        type: map
+        description: |
+            Groovy Map containing sample information
+            e.g. [ id:'test', single_end:false ]
+    - bam:
+        type: file
+        description: BAM/CRAM/SAM file
+        pattern: "*.{bam,cram,sam}"
+    - fasta:
+        type: file
+        description: FASTA reference file
+        pattern: "*.{fasta,fa}"
+output:
+    - meta:
+        type: map
+        description: |
+            Groovy Map containing sample information
+            e.g. [ id:'test', single_end:false ]
+    - mpileup:
+        type: file
+        description: mpileup file
+        pattern: "*.{mpileup}"
+    - version:
+        type: file
+        description: File containing software version
+        pattern: "*.{version.txt}"
+authors:
+    - "@drpatelh"
+    - "@joseespinosa"

software/samtools/view/functions.nf

@@ -0,0 +1,59 @@
+/*
+ * -----------------------------------------------------
+ *  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/samtools/view/main.nf

@@ -0,0 +1,34 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName } from './functions'
+
+params.options = [:]
+def options    = initOptions(params.options)
+
+process SAMTOOLS_VIEW {
+    tag "$meta.id"
+    publishDir "${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::samtools=1.10" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/samtools:1.10--h9402c20_2"
+    } else {
+        container "quay.io/biocontainers/samtools:1.10--h9402c20_2"
+    }
+
+    input:
+    tuple val(meta), path(bam)
+
+    output:
+    tuple val(meta), path("*.bam"), emit: bam
+    path  "*.version.txt"         , emit: version
+
+    script:
+    def software = getSoftwareName(task.process)
+    def prefix   = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    """
+    samtools view $options.args $bam > ${prefix}.bam
+    echo \$(samtools --version 2>&1) | sed 's/^.*samtools //; s/Using.*\$//' > ${software}.version.txt
+    """
+}

software/samtools/view/meta.yml

@@ -0,0 +1,64 @@
+name: samtools_view
+description: filter/convert SAM/BAM/CRAM file
+keywords:
+    - view
+    - bam
+    - sam
+    - cram
+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: hhttp://www.htslib.org/doc/samtools.html
+        doi: 10.1093/bioinformatics/btp352
+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:
+    - meta:
+        type: map
+        description: |
+            Groovy Map containing sample information
+            e.g. [ id:'test', single_end:false ]
+    - bam:
+        type: file
+        description: BAM/CRAM/SAM file
+        pattern: "*.{bam,cram,sam}"
+output:
+    - meta:
+        type: map
+        description: |
+            Groovy Map containing sample information
+            e.g. [ id:'test', single_end:false ]
+    - bam:
+        type: file
+        description: filtered/converted BAM/CRAM/SAM file
+        pattern: "*.{bam,cram,sam}"
+    - version:
+        type: file
+        description: File containing software version
+        pattern: "*.{version.txt}"
+authors:
+    - "@drpatelh"
+    - "@joseespinosa"

tests/software/samtools/main.nf

@@ -7,6 +7,8 @@ include { SAMTOOLS_IDXSTATS } from '../../../software/samtools/idxstats/main.nf'
 include { SAMTOOLS_INDEX } from '../../../software/samtools/index/main.nf' addParams( options: [:] )
 include { SAMTOOLS_SORT } from '../../../software/samtools/sort/main.nf' addParams( options: [:] )
 include { SAMTOOLS_STATS } from '../../../software/samtools/stats/main.nf' addParams( options: [:] )
+include { SAMTOOLS_VIEW } from '../../../software/samtools/view/main.nf' addParams( options: [:] )
+include { SAMTOOLS_MPILEUP } from '../../../software/samtools/mpileup/main.nf' addParams( options: [:] )
 
 workflow test_samtools_flagstat {
 
@@ -56,3 +58,23 @@ workflow test_samtools_stats {
 
     SAMTOOLS_STATS ( input )
 }
+
+workflow test_samtools_view {
+
+    def input = []
+    input = [ [ id:'test', single_end:false ], // meta map
+              file("${launchDir}/tests/data/bam/test.paired_end.sorted.bam", checkIfExists: true) ]
+
+    SAMTOOLS_VIEW ( input )
+}
+
+workflow test_samtools_mpileup {
+
+    def input = []
+    def fasta = []
+    input = [ [ id:'test', single_end:false ], // meta map
+              file("${launchDir}/tests/data/bam/test.paired_end.sorted.bam", checkIfExists: true) ]
+    fasta = [  file("${launchDir}/tests/data/fasta/E_coli/NC_010473.fa", checkIfExists: true)  ]
+
+    SAMTOOLS_MPILEUP ( input, fasta )
+}

tests/software/samtools/test.yml

@@ -42,3 +42,21 @@
   files:
     - path: output/samtools/test.paired_end.sorted.bam.stats
       md5sum: 06c183864d6e47ab89b0650cae831a93
+
+- name: Run samtools view test workflow
+  command: nextflow run ./tests/software/samtools/ -profile docker -entry test_samtools_view -c tests/config/nextflow.config
+  tags:
+    - samtools
+    - samtools_view
+  files:
+    - path: output/samtools/test.bam
+      md5sum: 7ea2c325e2249ed7ef44e1d9fdc4f8ff
+
+- name: Run samtools mpileup test workflow
+  command: nextflow run ./tests/software/samtools/ -profile docker -entry test_samtools_mpileup -c tests/config/nextflow.config
+  tags:
+    - samtools
+    - samtools_mpileup
+  files:
+    - path: output/samtools/test.mpileup
+      md5sum: 95c2646fa7bd535207722008c581c81d