add rapidnj module (#421)

* add rapidnj module

* remove trailing whitespace

* Apply suggestions from code review

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

software/rapidnj/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/rapidnj/main.nf

@@ -0,0 +1,44 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process RAPIDNJ {
+    label 'process_medium'
+    publishDir "${params.outdir}",
+        mode: params.publish_dir_mode,
+        saveAs: { filename -> saveFiles(filename:filename, options:params.options, publish_dir:getSoftwareName(task.process), publish_id:'') }
+
+    conda (params.enable_conda ? "bioconda::rapidnj=2.3.2 conda-forge::biopython=1.78" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/mulled-v2-805c6e0f138f952f9c61cdd57c632a1a263ea990:3c52e4c8da6b3e4d69b9ca83fa4d366168898179-0"
+    } else {
+        container "quay.io/biocontainers/mulled-v2-805c6e0f138f952f9c61cdd57c632a1a263ea990:3c52e4c8da6b3e4d69b9ca83fa4d366168898179-0"
+    }
+
+    input:
+    path alignment
+
+    output:
+    path "*.sth"        , emit: stockholm_alignment
+    path "*.tre"        , emit: phylogeny
+    path "*.version.txt", emit: version
+
+    script:
+    def software = getSoftwareName(task.process)
+    """
+    python \\
+        -c 'from Bio import SeqIO; SeqIO.convert("$alignment", "fasta", "alignment.sth", "stockholm")'
+
+    rapidnj \\
+        alignment.sth \\
+        $options.args \\
+        -i sth \\
+        -c $task.cpus \\
+        -x rapidnj_phylogeny.tre
+
+    # Doesn't appear to be a way of getting the version number
+    echo 2.3.2 > ${software}.version.txt
+    """
+}

software/rapidnj/meta.yml

@@ -0,0 +1,37 @@
+name: rapidnj
+description: Produces a Newick format phylogeny from a multiple sequence alignment using a Neighbour-Joining algorithm. Capable of bacterial genome size alignments.
+keywords:
+  - phylogeny
+  - newick
+  - neighbour-joining
+tools:
+  - rapidnj:
+      description: RapidNJ is an algorithmic engineered implementation of canonical neighbour-joining. It uses an efficient search heuristic to speed-up the core computations of the neighbour-joining method that enables RapidNJ to outperform other state-of-the-art neighbour-joining implementations.
+      homepage: https://birc.au.dk/software/rapidnj
+      documentation: https://birc.au.dk/software/rapidnj
+      tool_dev_url: https://github.com/somme89/rapidNJ
+      doi: "doi:10.1007/978-3-540-87361-7_10"
+      licence: ['GPL v2']
+
+input:
+  - alignment:
+      type: file
+      description: A FASTA format multiple sequence alignment file
+      pattern: "*.{fasta,fas,fa,mfa}"
+
+output:
+  - version:
+      type: file
+      description: File containing software version
+      pattern: "*.{version.txt}"
+  - phylogeny:
+      type: file
+      description: A phylogeny in Newick format
+      pattern: "*.{tre}"
+  - stockholm_alignment:
+      type: file
+      description: An alignment in Stockholm format
+      pattern: "*.{sth}"
+authors:
+  - "@aunderwo"
+  - "@avantonder"

tests/config/pytest_software.yml

@@ -108,15 +108,6 @@ blast/makeblastdb:
   - software/blast/makeblastdb/**
   - tests/software/blast/makeblastdb/**
 
-bowtie2/align:
-  - software/bowtie2/align/**
-  - software/bowtie2/build/**
-  - tests/software/bowtie2/align/**
-
-bowtie2/build:
-  - software/bowtie2/build/**
-  - tests/software/bowtie2/build_test/**
-
 bowtie/align:
   - software/bowtie/align/**
   - software/bowtie/build/**
@@ -126,6 +117,15 @@ bowtie/build:
   - software/bowtie/build/**
   - tests/software/bowtie/build_test/**
 
+bowtie2/align:
+  - software/bowtie2/align/**
+  - software/bowtie2/build/**
+  - tests/software/bowtie2/align/**
+
+bowtie2/build:
+  - software/bowtie2/build/**
+  - tests/software/bowtie2/build_test/**
+
 bwa/index:
   - software/bwa/index/**
   - tests/software/bwa/index/**
@@ -386,6 +386,10 @@ quast:
   - software/quast/**
   - tests/software/quast/**
 
+rapidnj:
+  - software/rapidnj/**
+  - tests/software/rapidnj/**
+
 salmon/index:
   - software/salmon/index/**
   - tests/software/salmon/index/**

tests/software/rapidnj/main.nf

@@ -0,0 +1,12 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { RAPIDNJ } from '../../../software/rapidnj/main.nf' addParams( options: [:] )
+
+workflow test_rapidnj {
+    
+    input = [ file(params.test_data['sarscov2']['genome']['informative_sites_fas'], checkIfExists: true) ]
+
+    RAPIDNJ ( input )
+}

tests/software/rapidnj/test.yml

@@ -0,0 +1,9 @@
+- name: rapidnj
+  command: nextflow run ./tests/software/rapidnj -entry test_rapidnj -c tests/config/nextflow.config
+  tags:
+    - rapidnj
+  files:
+    - path: output/rapidnj/alignment.sth
+      md5sum: d2e995c5dd3e3a8212a98414ae5b5de7
+    - path: output/rapidnj/rapidnj_phylogeny.tre
+      md5sum: 775909ea40138101976592cfa1814a1d