Adding minia for viralrecon (#267)

* initial 'modules create' of minia

* fixed tests

* finished meta.yml

* fixed filters.yml

* resolved issues in pytest_software.yml

* add newline

* Update software/minia/main.nf

Co-authored-by: Harshil Patel <drpatelh@users.noreply.github.com>

software/minia/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/minia/main.nf

@@ -0,0 +1,41 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName } from './functions'
+
+params.options = [:]
+def options    = initOptions(params.options)
+process MINIA {
+    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::minia=3.2.4" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/minia:3.2.4--he513fc3_0"
+    } else {
+        container "quay.io/biocontainers/minia:3.2.4--he513fc3_0"
+    }
+
+    input:
+    tuple val(meta), path(reads)
+
+    output:
+    tuple val(meta), path('*.contigs.fa'), emit: contigs
+    tuple val(meta), path('*.unitigs.fa'), emit: unitigs
+    tuple val(meta), path('*.h5')        , emit: h5
+    path  '*.version.txt'                , emit: version
+
+    script:
+    def software = getSoftwareName(task.process)
+    def prefix   = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    """
+    echo "${reads.join("\n")}" > input_files.txt
+    minia \\
+        $options.args \\
+        -nb-cores $task.cpus \\
+        -in input_files.txt \\
+        -out $prefix
+    echo \$(minia --version 2>&1) | sed 's/^.*Minia version //; s/ .*\$//' > ${software}.version.txt
+    """
+}

software/minia/meta.yml

@@ -0,0 +1,67 @@
+name: minia
+description: Minia is a short-read assembler based on a de Bruijn graph
+keywords:
+  - assembly
+tools:
+  - minia:
+      description: |
+        Minia is a short-read assembler based on a de Bruijn graph, capable of assembling
+        a human genome on a desktop computer in a day. The output of Minia is a set of contigs.
+      homepage: https://github.com/GATB/minia
+      documentation: https://github.com/GATB/minia
+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 ]
+  - reads:
+      type: file
+      description: Input reads in FastQ format
+      pattern: "*.{fastq.gz, fastq}"
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - contigs:
+      type: file
+      description: The assembled contigs
+      pattern: "*.contigs.fa"
+  - unitigs:
+      type: file
+      description: The assembled unitigs
+      pattern: "*.unitigs.fa"
+  - h5:
+      type: file
+      description: Minia output h5 file
+      pattern: "*{.h5}"
+  - version:
+      type: file
+      description: File containing software version
+      pattern: "*.{version.txt}"
+authors:
+  - "@drpatelh"
+  - "@kevinmenden"

tests/config/pytest_software.yml

@@ -66,10 +66,6 @@ bismark_genome_preparation:
   - software/bismark/genome_preparation/**
   - tests/software/bismark/genome_preparation/**
 
-blast_makeblastdb:
-  - software/blast/makeblastdb/**
-  - tests/software/blast/makeblastdb/**
-
 blast_blastn:
   - software/blast/blastn/**
   - tests/software/blast/blastn/**
@@ -192,6 +188,10 @@ methyldackel_mbias:
   - software/methyldackel/mbias/**
   - tests/software/methyldackel/mbias/**
 
+minia:
+  - software/minia/**
+  - tests/software/minia/**
+
 minimap2_align:
   - software/minimap2/align/**
   - tests/software/minimap2/align/**
@@ -313,14 +313,14 @@ trimgalore:
   - software/trimgalore/**
   - tests/software/trimgalore/**
 
-untar:
-  - software/untar/**
-  - tests/software/untar/**
-
 ucsc_bedgraphtobigwig:
   - software/ucsc/bedgraphtobigwig/**
   - tests/software/ucsc/bedgraphtobigwig/**
 
+untar:
+  - software/untar/**
+  - tests/software/untar/**
+
 gatk4_variantfiltration:
   - software/gatk4/variantfiltration/**
   - tests/software/gatk4/variantfiltration/**

tests/software/minia/main.nf

@@ -0,0 +1,15 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { MINIA } from '../../../software/minia/main.nf' addParams( options: [:] )
+
+workflow test_minia {
+
+    def input = []
+    input = [ [ id:'test' ], // meta map
+              [file("${launchDir}/tests/data/genomics/sarscov2/fastq/test_1.fastq.gz", checkIfExists: true),
+              file("${launchDir}/tests/data/genomics/sarscov2/fastq/test_2.fastq.gz", checkIfExists: true)] ]
+
+    MINIA ( input )
+}

tests/software/minia/test.yml

@@ -0,0 +1,8 @@
+- name: Run tests for minia - test_minia
+  command: nextflow run tests/software/minia -entry test_minia -c tests/config/nextflow.config
+  tags:
+    - minia
+  files:
+    - path: output/minia/test.h5
+    - path: output/minia/test.contigs.fa
+    - path: output/minia/test.unitigs.fa