chores: add gatk/createsequencedictionary module

This commit is contained in:
MaxUlysse 2021-02-16 13:40:06 +01:00
parent 8c4d201dd5
commit 7f35a1742d
6 changed files with 169 additions and 0 deletions

4
.github/filters.yml vendored
View file

@ -128,6 +128,10 @@ fastqc:
- software/fastqc/** - software/fastqc/**
- tests/software/fastqc/** - tests/software/fastqc/**
gatk_createsequencedictionary:
- software/gatk/createsequencedictionary/**
- tests/software/gatk/createsequencedictionary/**
gffread: gffread:
- software/gffread/** - software/gffread/**
- tests/software/gffread/** - tests/software/gffread/**

View file

@ -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"
}
}
}

View file

@ -0,0 +1,38 @@
// Import generic module functions
include { initOptions; saveFiles; getSoftwareName } from './functions'
params.options = [:]
def options = initOptions(params.options)
process GATK_CREATESEQUENCEDICTIONARY {
tag "$fasta"
label 'process_medium'
publishDir "${params.outdir}",
mode: params.publish_dir_mode,
saveAs: { filename -> saveFiles(filename:filename, options:params.options, publish_dir:getSoftwareName(task.process), publish_id:'') }
conda (params.enable_conda ? "bioconda::gatk4=4.1.9.0" : null)
if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
container "https://depot.galaxyproject.org/singularity/gatk4:4.1.9.0--py39_0"
} else {
container "quay.io/biocontainers/gatk4:4.1.9.0--py39_0"
}
input:
path fasta
output:
path "*.dict" , emit: dict
path "*.version.txt" , emit: version
script:
def software = getSoftwareName(task.process)
"""
gatk --java-options "-Xmx${task.memory.toGiga()}g" \\
CreateSequenceDictionary \\
--REFERENCE $fasta \\
$options.args
echo \$(gatk CreateSequenceDictionary --version 2>&1) | sed 's/^.*(GATK) v//; s/ HTSJDK.*\$//' > ${software}.version.txt
"""
}

View file

@ -0,0 +1,51 @@
name: gatk_createsequencedictionary
description: Creates a sequence dictionary for a reference sequence
keywords:
- dictionary
- fasta
tools:
- gatk:
description: |
Developed in the Data Sciences Platform at the Broad Institute, the toolkit offers a wide variety of tools
with a primary focus on variant discovery and genotyping. Its powerful processing engine
and high-performance computing features make it capable of taking on projects of any size.
homepage: https://gatk.broadinstitute.org/hc/en-us
documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
doi: 10.1158/1538-7445.AM2017-3590
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 fasta file
pattern: "*.{fasta,fa}"
output:
- bam:
type: file
description: gatk dictionary file
pattern: "*.{dict}"
- version:
type: file
description: File containing software version
pattern: "*.{version.txt}"
authors:
- "@maxulysse"

View file

@ -0,0 +1,9 @@
#!/usr/bin/env nextflow
nextflow.enable.dsl = 2
include { GATK_CREATESEQUENCEDICTIONARY } from '../../../../software/gatk/createsequencedictionary/main.nf' addParams( options: [:] )
workflow test_gatk_createsequencedictionary {
GATK_CREATESEQUENCEDICTIONARY ( file("${launchDir}/tests/data/fasta/E_coli/NC_010473.fa", checkIfExists: true) )
}

View file

@ -0,0 +1,8 @@
- name: gatk createsequencedictionary
command: nextflow run ./tests/software/gatk/createsequencedictionary -entry test_gatk_createsequencedictionary -c tests/config/nextflow.config
tags:
- gatk
- gatk_createsequencedictionary
files:
- path: output/gatk/NC_010473.dict
md5sum: 2b5e6b5d82553ddcaf85156ba856a55c