Merge pull request #218 from phue/samtools_faidx

add samtools/faidx + tests

.github/filters.yml

@@ -193,6 +193,10 @@ salmon_quant:
   - software/salmon/quant/**
   - tests/software/salmon/quant/**
 
+samtools_faidx:
+  - software/samtools/faidx/**
+  - tests/software/samtools/faidx/**
+
 samtools_flagstat:
   - software/samtools/flagstat/**
   - tests/software/samtools/flagstat/**

software/samtools/faidx/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/faidx/main.nf

@@ -0,0 +1,33 @@
+// Import generic module functions
+include { saveFiles; getSoftwareName } from './functions'
+
+params.options = [:]
+
+process SAMTOOLS_FAIDX {
+    tag "$fasta"
+    label 'process_low'
+    publishDir "${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::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:
+    path fasta
+
+    output:
+    path "*.fai"        , emit: fai
+    path "*.version.txt", emit: version
+
+    script:
+    def software = getSoftwareName(task.process)
+    """
+    samtools faidx $fasta
+    echo \$(samtools --version 2>&1) | sed 's/^.*samtools //; s/Using.*\$//' > ${software}.version.txt
+    """
+}

software/samtools/faidx/meta.yml

@@ -0,0 +1,53 @@
+name: samtools_faidx
+description: Index FASTA file
+keywords:
+    - index
+    - fasta
+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
+        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:
+    - fasta:
+        type: file
+        description: FASTA file
+        pattern: "*.{fa,fasta}"
+output:
+    - fai:
+        type: file
+        description: FASTA index file
+        pattern: "*.{fai}"
+    - version:
+        type: file
+        description: File containing software version
+        pattern: "*.{version.txt}"
+authors:
+    - "@drpatelh"
+    - "@ewels"
+    - "@phue"

tests/software/samtools/faidx/main.nf

@@ -0,0 +1,10 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { SAMTOOLS_FAIDX } from '../../../../software/samtools/faidx/main.nf' addParams( options: [:] )
+
+workflow test_samtools_faidx {
+
+    SAMTOOLS_FAIDX ( file("${launchDir}/tests/data/fasta/E_coli/NC_010473.fa", checkIfExists: true) )
+}

tests/software/samtools/faidx/test.yml

@@ -0,0 +1,8 @@
+- name: Run samtools faidx test workflow
+  command: nextflow run ./tests/software/samtools/faidx -entry test_samtools_faidx -c tests/config/nextflow.config
+  tags:
+    - samtools
+    - samtools_faidx
+  files:
+    - path: output/samtools/NC_010473.fa.fai
+      md5sum: 082de78397fbe9bd3500c7d5c9edc03e