Merge branch 'get-chrom-sizes' into dev (cfcff692) · Commits · Chaos / Cutandrun

modules.json

+3 −0

Original line number	Diff line number	Diff line
		@@ -27,6 +27,9 @@
		"cat/fastq": {
		"git_sha": "3aacd46da2b221ed47aaa05c413a828538d2c2ae"
		},
		"custom/getchromsizes": {
		"git_sha": "13b8a16f4a6945af9df146b67972eb70b52e9844"
		},
		"deeptools/computematrix": {
		"git_sha": "49da8642876ae4d91128168cd0db4f1c858d7792"
		},

modules/nf-core/modules/custom/getchromsizes/functions.nf

0 → 100644

+78 −0

Original line number	Diff line number	Diff line
		//
		// 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		// Do not publish versions.yml unless running from pytest workflow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/local/get_chrom_sizes.nf→modules/nf-core/modules/custom/getchromsizes/main.nf

+4 −8

Original line number	Diff line number	Diff line
		// Import generic module functions
		include { initOptions; saveFiles; getSoftwareName; getProcessName } from './common/functions'
		include { initOptions; saveFiles; getSoftwareName; getProcessName } from './functions'

		params.options = [:]
		options = initOptions(params.options)

		/*
		* Get chromosome sizes from a fasta file
		*/
		process GET_CHROM_SIZES {
		process CUSTOM_GETCHROMSIZES {
		tag "$fasta"
		label 'process_low'
		publishDir "${params.outdir}",
		mode: params.publish_dir_mode,
		saveAs: { filename -> saveFiles(filename:filename, options:params.options, publish_dir:"genome", publish_id:'') }
		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) {
		@@ -30,14 +27,13 @@ process GET_CHROM_SIZES {
		path "versions.yml", emit: versions

		script:
		def software = 'samtools'
		"""
		samtools faidx $fasta
		cut -f 1,2 ${fasta}.fai > ${fasta}.sizes

		cat <<-END_VERSIONS > versions.yml
		${getProcessName(task.process)}:
		samtools: \$(echo \$(samtools --version 2>&1) \| sed 's/^.samtools //; s/Using.\$//')
		${getSoftwareName(task.process)}: \$(echo \$(samtools --version 2>&1) \| sed 's/^.samtools //; s/Using.\$//')
		END_VERSIONS
		"""
		}

modules/nf-core/modules/custom/getchromsizes/meta.yml

0 → 100644

+39 −0

Original line number	Diff line number	Diff line
		name: custom_getchromsizes
		description: Generates a FASTA file of chromosome sizes and a fasta index file
		keywords:
		- fasta
		- chromosome
		- indexing
		tools:
		- samtools:
		description: Tools for dealing with SAM, BAM and CRAM files
		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:
		- fasta:
		type: file
		description: FASTA file
		pattern: "*.{fasta}"

		output:
		- sizes:
		type: file
		description: File containing chromosome lengths
		pattern: "*.{sizes}"
		- fai:
		type: file
		description: FASTA index file
		pattern: "*.{fai}"
		- versions:
		type: file
		description: File containing software version
		pattern: "versions.yml"


		authors:
		- "@tamara-hodgetts"
		- "@chris-cheshire"

subworkflows/local/prepare_genome.nf

+4 −4

Original line number	Diff line number	Diff line
		@@ -11,8 +11,8 @@ include { GUNZIP as GUNZIP_FASTA } from "../../modules/nf-co
		include { GUNZIP as GUNZIP_SPIKEIN_FASTA } from "../../modules/nf-core/modules/gunzip/main.nf" addParams( options: params.spikein_genome_options )
		include { GUNZIP as GUNZIP_GTF } from "../../modules/nf-core/modules/gunzip/main.nf" addParams( options: params.genome_options )
		include { GUNZIP as GUNZIP_BED } from "../../modules/nf-core/modules/gunzip/main.nf" addParams( options: params.genome_options )
		include { GET_CHROM_SIZES } from "../../modules/local/get_chrom_sizes" addParams( options: params.genome_options )
		include { GET_CHROM_SIZES as GET_SPIKEIN_CHROM_SIZES } from "../../modules/local/get_chrom_sizes" addParams( options: params.spikein_genome_options )
		include { CUSTOM_GETCHROMSIZES } from "../../modules/nf-core/modules/custom/getchromsizes/main.nf" addParams( options: params.genome_options )
		include { CUSTOM_GETCHROMSIZES as GET_SPIKEIN_CHROM_SIZES } from "../../modules/nf-core/modules/custom/getchromsizes/main.nf" addParams( options: params.spikein_genome_options )
		include { UNTAR as UNTAR_BT2_INDEX } from "../../modules/nf-core/modules/untar/main.nf" addParams( options: params.bt2_index_options )
		include { UNTAR as UNTAR_SPIKEIN_BT2_INDEX } from "../../modules/nf-core/modules/untar/main.nf" addParams( options: params.bt2_spikein_index_options )
		include { BOWTIE2_BUILD } from "../../modules/nf-core/modules/bowtie2/build/main" addParams( options: params.bt2_index_options )
		@@ -72,8 +72,8 @@ workflow PREPARE_GENOME {
		/*
		* Create chromosome sizes file
		*/
		ch_chrom_sizes = GET_CHROM_SIZES ( ch_fasta ).sizes
		ch_versions = ch_versions.mix(GET_CHROM_SIZES.out.versions)
		ch_chrom_sizes = CUSTOM_GETCHROMSIZES ( ch_fasta ).sizes
		ch_versions = ch_versions.mix(CUSTOM_GETCHROMSIZES.out.versions)


		/*

Admin message