Prodigal (#333)

* WIP add prodigal

* Implement module and add tests

* Fix EC lint

software/prodigal/functions.nf

@@ -0,0 +1,60 @@
+/*
+ * -----------------------------------------------------
+ *  Utility functions used in nf-core DSL2 module files
+ * -----------------------------------------------------
+ */
+
+/*
+ * Extract name of software tool from process name using $task.process
+ */
+def getSoftwareName(task_process) {
+    return task_process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()
+}
+
+/*
+ * Function to initialise default values and to generate a Groovy Map of available options for nf-core modules
+ */
+def initOptions(Map args) {
+    def Map options = [:]
+    options.args          = args.args ?: ''
+    options.args2         = args.args2 ?: ''
+    options.args3         = args.args3 ?: ''
+    options.publish_by_id = args.publish_by_id ?: false
+    options.publish_dir   = args.publish_dir ?: ''
+    options.publish_files = args.publish_files
+    options.suffix        = args.suffix ?: ''
+    return options
+}
+
+/*
+ * Tidy up and join elements of a list to return a path string
+ */
+def getPathFromList(path_list) {
+    def paths = path_list.findAll { item -> !item?.trim().isEmpty() }  // Remove empty entries
+    paths = paths.collect { it.trim().replaceAll("^[/]+|[/]+\$", "") } // Trim whitespace and trailing slashes
+    return paths.join('/')
+}
+
+/*
+ * Function to save/publish module results
+ */
+def saveFiles(Map args) {
+    if (!args.filename.endsWith('.version.txt')) {
+        def ioptions = initOptions(args.options)
+        def path_list = [ ioptions.publish_dir ?: args.publish_dir ]
+        if (ioptions.publish_by_id) {
+            path_list.add(args.publish_id)
+        }
+        if (ioptions.publish_files instanceof Map) {
+            for (ext in ioptions.publish_files) {
+                if (args.filename.endsWith(ext.key)) {
+                    def ext_list = path_list.collect()
+                    ext_list.add(ext.value)
+                    return "${getPathFromList(ext_list)}/$args.filename"
+                }
+            }
+        } else if (ioptions.publish_files == null) {
+            return "${getPathFromList(path_list)}/$args.filename"
+        }
+    }
+}

software/prodigal/main.nf

@@ -0,0 +1,46 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process PRODIGAL {
+    tag "$meta.id"
+    label 'process_low'
+    publishDir "${params.outdir}",
+        mode: params.publish_dir_mode,
+        saveAs: { filename -> saveFiles(filename:filename, options:params.options, publish_dir:getSoftwareName(task.process), publish_id:meta.id) }
+
+    conda (params.enable_conda ? "bioconda::prodigal=2.6.3" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/prodigal:2.6.3--h516909a_2"
+    } else {
+        container "quay.io/biocontainers/prodigal:2.6.3--h516909a_2"
+    }
+
+    input:
+    tuple val(meta), path(genome)
+    val(output_format)
+
+    output:
+    tuple val(meta), path("${prefix}.${output_format}"), emit: gene_annotations
+    tuple val(meta), path("${prefix}.fna"), emit: nucleotide_fasta
+    tuple val(meta), path("${prefix}.faa"), emit: amino_acid_fasta
+    tuple val(meta), path("${prefix}_all.txt"), emit: all_gene_annotations
+    path "*.version.txt"          , emit: version
+
+    script:
+    def software = getSoftwareName(task.process)
+    prefix = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    """
+    prodigal -i "${genome}" \\
+        $options.args \\
+        -f $output_format \\
+        -d "${prefix}.fna" \\
+        -o "${prefix}.${output_format}" \\
+        -a "${prefix}.faa" \\
+        -s "${prefix}_all.txt"
+
+    echo \$(prodigal -v 2>&1) | sed -n 's/Prodigal V\\(.*\\):.*/\\1/p' > ${software}.version.txt
+    """
+}

software/prodigal/meta.yml

@@ -0,0 +1,47 @@
+name: prodigal
+## TODO nf-core: Add a description of the module and list keywords
+description: write your description here
+keywords:
+  - sort
+tools:
+  - prodigal:
+      ## TODO nf-core: Add a description and other details for the software below
+      description: Prodigal (Prokaryotic Dynamic Programming Genefinding Algorithm) is a microbial (bacterial and archaeal) gene finding program
+      homepage: {}
+      documentation: {}
+      tool_dev_url: {}
+      doi: ""
+      licence: ['GPL v3']
+
+## TODO nf-core: Add a description of all of the variables used as input
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  ## TODO nf-core: Delete / customise this example input
+  - bam:
+      type: file
+      description: BAM/CRAM/SAM file
+      pattern: "*.{bam,cram,sam}"
+
+## TODO nf-core: Add a description of all of the variables used as output
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - version:
+      type: file
+      description: File containing software version
+      pattern: "*.{version.txt}"
+  ## TODO nf-core: Delete / customise this example output
+  - bam:
+      type: file
+      description: Sorted BAM/CRAM/SAM file
+      pattern: "*.{bam,cram,sam}"
+
+authors:
+  - "@grst"

tests/config/pytest_software.yml

@@ -299,6 +299,10 @@ preseq_lcextrap:
   - software/preseq/lcextrap/**
   - tests/software/preseq/lcextrap/**
 
+prodigal:
+  - software/prodigal/**
+  - tests/software/prodigal/**
+
 prokka:
   - software/prokka/**
   - tests/software/prokka/**

tests/software/prodigal/main.nf

@@ -0,0 +1,14 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { PRODIGAL } from '../../../software/prodigal/main.nf' addParams( options: [:] )
+
+workflow test_prodigal {
+
+    def input = []
+    input = [ [ id:'test' ], // meta map
+              file("${launchDir}/tests/data/genomics/sarscov2/fasta/test_genome.fasta", checkIfExists: true) ]
+
+    PRODIGAL ( input , "gff")
+}

tests/software/prodigal/test.yml

@@ -0,0 +1,13 @@
+- name: prodigal test_prodigal
+  command: nextflow run tests/software/prodigal -entry test_prodigal -c tests/config/nextflow.config
+  tags:
+    - prodigal
+  files:
+    - path: output/prodigal/test.fna
+      md5sum: 1bc8a05bcb72a3c324f5e4ffaa716d3b
+    - path: output/prodigal/test.gff
+      md5sum: 612c2724c2891c63350f171f74165757
+    - path: output/prodigal/test.faa
+      md5sum: 7168b854103f3586ccfdb71a44c389f7
+    - path: output/prodigal/test_all.txt
+      md5sum: e6d6c50f0c39e5169f84ae3c90837fa9