Add MAFFT module (#1351)

Signed-off-by: Thomas A. Christensen II <25492070+MillironX@users.noreply.github.com>

Co-authored-by: Sateesh <33637490+sateeshperi@users.noreply.github.com>
Co-authored-by: Robert A. Petit III <robbie.petit@gmail.com>

modules/mafft/main.nf

@@ -0,0 +1,35 @@
+process MAFFT {
+    tag "$meta.id"
+    label 'process_high'
+
+    conda (params.enable_conda ? "bioconda::mafft=7.490" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/mafft:7.490--h779adbc_0':
+        'quay.io/biocontainers/mafft:7.490--h779adbc_0' }"
+
+    input:
+    tuple val(meta), path(fasta)
+
+    output:
+    tuple val(meta), path("*.fas"), emit: fas
+    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}"
+    """
+    mafft \\
+        --thread ${task.cpus} \\
+        ${args} \\
+        ${fasta} \\
+        > ${prefix}.fas
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        mafft: \$(mafft --version 2>&1 | sed 's/^v//' | sed 's/ (.*)//')
+    END_VERSIONS
+    """
+}

modules/mafft/meta.yml

@@ -0,0 +1,42 @@
+name: mafft
+description: Multiple sequence alignment using MAFFT
+keywords:
+  - msa
+  - multiple sequence alignment
+tools:
+  - mafft:
+      description: Multiple alignment program for amino acid or nucleotide sequences based on fast Fourier transform
+      homepage: https://mafft.cbrc.jp/alignment/software/
+      documentation: https://mafft.cbrc.jp/alignment/software/manual/manual.html
+      tool_dev_url: https://mafft.cbrc.jp/alignment/software/source.html
+      doi: "10.1093/nar/gkf436"
+      licence: ['BSD']
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - fasta:
+      type: file
+      description: FASTA file containing the sequences to align
+      pattern: "*.{fa,fasta}"
+
+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"
+  - fas:
+      type: file
+      description: Aligned sequences in FASTA format
+      pattern: "*.{fas}"
+
+authors:
+  - "@MillironX"

tests/config/pytest_modules.yml

@@ -964,6 +964,10 @@ macs2/callpeak:
   - modules/macs2/callpeak/**
   - tests/modules/macs2/callpeak/**
 
+mafft:
+  - modules/mafft/**
+  - tests/modules/mafft/**
+
 malt/build:
   - modules/malt/build/**
   - tests/modules/malt/build_test/**

tests/modules/mafft/main.nf

@@ -0,0 +1,15 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { MAFFT } from '../../../modules/mafft/main.nf'
+
+workflow test_mafft {
+
+    input = [
+        [ id:'test', single_end:false ], // meta map
+        file(params.test_data['sarscov2']['illumina']['scaffolds_fasta'], checkIfExists: true)
+    ]
+
+    MAFFT ( input )
+}

tests/modules/mafft/nextflow.config

@@ -0,0 +1,6 @@
+process {
+
+    publishDir = { "${params.outdir}/${task.process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()}" }
+    ext.args = "--auto"
+
+}

tests/modules/mafft/test.yml

@@ -0,0 +1,9 @@
+- name: mafft test_mafft
+  command: nextflow run tests/modules/mafft -entry test_mafft -c tests/config/nextflow.config
+  tags:
+    - mafft
+  files:
+    - path: output/mafft/test.fas
+      md5sum: 23426611f4a0df532b6708f072bd445b
+    - path: output/mafft/versions.yml
+      md5sum: b1b5ab3728ae17401808335f1c8f8215