add mobsuite_recon module (#1270)

* add mobsuite_recon module

* Update main.nf

* Update nextflow.config

* Update test.yml

Co-authored-by: Sateesh <33637490+sateeshperi@users.noreply.github.com>

modules/mobsuite/recon/main.nf

@@ -0,0 +1,45 @@
+process MOBSUITE_RECON {
+    tag "$meta.id"
+    label 'process_medium'
+
+    conda (params.enable_conda ? "bioconda::mob_suite=3.0.3" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/mob_suite%3A3.0.3--pyhdfd78af_0':
+        'quay.io/biocontainers/mob_suite:3.0.3--pyhdfd78af_0' }"
+
+    input:
+    tuple val(meta), path(fasta)
+
+    output:
+    tuple val(meta), path("results/chromosome.fasta")    , emit: chromosome
+    tuple val(meta), path("results/contig_report.txt")   , emit: contig_report
+    tuple val(meta), path("results/plasmid_*.fasta")     , emit: plasmids        , optional: true
+    tuple val(meta), path("results/mobtyper_results.txt"), emit: mobtyper_results, optional: true
+    path "versions.yml"                                  , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    def is_compressed = fasta.getName().endsWith(".gz") ? true : false
+    def fasta_name = fasta.getName().replace(".gz", "")
+    """
+    if [ "$is_compressed" == "true" ]; then
+        gzip -c -d $fasta > $fasta_name
+    fi
+
+    mob_recon \\
+        --infile $fasta_name \\
+        $args \\
+        --num_threads $task.cpus \\
+        --outdir results \\
+        --sample_id $prefix
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        mobsuite: \$(echo \$(mob_recon --version 2>&1) | sed 's/^.*mob_recon //; s/ .*\$//')
+    END_VERSIONS
+    """
+}

modules/mobsuite/recon/meta.yml

@@ -0,0 +1,54 @@
+name: mobsuite_recon
+description: A tool to reconstruct plasmids in bacterial assemblies
+keywords:
+  - bacteria
+  - plasmid
+tools:
+  - mobsuite:
+      description: Software tools for clustering, reconstruction and typing of plasmids from draft assemblies.
+      homepage: https://github.com/phac-nml/mob-suite
+      documentation: https://github.com/phac-nml/mob-suite
+      tool_dev_url: https://github.com/phac-nml/mob-suite
+      doi: "10.1099/mgen.0.000435"
+      licence: ['Apache License, Version 2.0']
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - fasta:
+      type: file
+      description: A bacterial genome assembly in FASTA format
+      pattern: "*.{fasta,fa,fna}"
+
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+  - chromosome:
+      type: file
+      description: FASTA file of all contigs found to belong to the chromosome
+      pattern: "chromosome.fasta"
+  - contig_report:
+      type: file
+      description: Assignment of the contig to chromosome or a particular plasmid grouping
+      pattern: "contig_report.txt"
+  - plasmids:
+      type: file
+      description: Each plasmid group is written to an individual FASTA
+      pattern: "plasmid_*.fasta"
+  - mobtyper_results:
+      type: file
+      description: Aggregate MOB-typer report files for all identified plasmid
+      pattern: "mobtyper_results.txt"
+
+authors:
+  - "@rpetit3"

tests/config/pytest_modules.yml

@@ -1024,6 +1024,10 @@ mlst:
   - modules/mlst/**
   - tests/modules/mlst/**
 
+mobsuite/recon:
+  - modules/mobsuite/recon/**
+  - tests/modules/mobsuite/recon/**
+
 mosdepth:
   - modules/mosdepth/**
   - tests/modules/mosdepth/**

tests/modules/mobsuite/recon/main.nf

@@ -0,0 +1,13 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { MOBSUITE_RECON } from '../../../../modules/mobsuite/recon/main.nf'
+
+workflow test_mobsuite_recon {
+
+    input = [ [ id:'test', single_end:false ], // meta map
+              file(params.test_data['sarscov2']['genome']['genome_fasta'], checkIfExists: true) ]
+
+    MOBSUITE_RECON ( input )
+}

tests/modules/mobsuite/recon/nextflow.config

@@ -0,0 +1,5 @@
+process {
+
+    publishDir = { "${params.outdir}/${task.process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()}" }
+    
+}

tests/modules/mobsuite/recon/test.yml

@@ -0,0 +1,12 @@
+- name: mobsuite recon test_mobsuite_recon
+  command: nextflow run tests/modules/mobsuite/recon -entry test_mobsuite_recon -c tests/config/nextflow.config
+  tags:
+    - mobsuite/recon
+    - mobsuite
+  files:
+    - path: output/mobsuite/results/chromosome.fasta
+      md5sum: 33b2a0fa321c73c6ba8d8272dd53c6d4
+    - path: output/mobsuite/results/contig_report.txt
+      md5sum: a0ae364a9f2b475f77588d0b3c24b857
+    - path: output/mobsuite/versions.yml
+      md5sum: 7f7a0f8957394b0e526233a0edb8e20a