Add vserach usearch_global

modules/vsearch/usearchglobal/main.nf

@@ -0,0 +1,51 @@
+// TODO nf-core: A module file SHOULD only define input and output files as command-line parameters.
+//               All other parameters MUST be provided using the "task.ext" directive, see here:
+//               https://www.nextflow.io/docs/latest/process.html#ext
+//               where "task.ext" is a string.
+//               Any parameters that need to be evaluated in the context of a particular sample
+//               e.g. single-end/paired-end data MUST also be defined and evaluated appropriately.
+// TODO nf-core: Software that can be piped together SHOULD be added to separate module files
+//               unless there is a run-time, storage advantage in implementing in this way
+//               e.g. it's ok to have a single module for bwa to output BAM instead of SAM:
+//                 bwa mem | samtools view -B -T ref.fasta
+
+process VSEARCH_USEARCHGLOBAL {
+    tag '$queryfasta'
+    label 'process_low'
+
+    conda (params.enable_conda ? "bioconda::vsearch=2.21.1" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/vsearch:2.21.1--hf1761c0_1':
+        'quay.io/biocontainers/vsearch:2.21.1--h95f258a_0' }"
+
+    input:
+    path queryfasta
+    path db
+    val  outprefix
+
+    // TODO nf-core: Where applicable please provide/convert compressed files as input/output
+    //               e.g. "*.fastq.gz" and NOT "*.fastq", "*.bam" and NOT "*.sam" etc.
+
+    output:
+    path ("*.tsv")        , emit: tsv
+    path "versions.yml"   , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    """
+    vsearch \\
+        --usearch_global $queryfasta \\
+        --db $db \\
+        --threads $task.cpus \\
+        $args \\
+        --blast6out ${outprefix}.tsv
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        vsearch: \$(vsearch --version 2>&1 | head -n 1 | sed 's/vsearch //g' | sed 's/,.*//g' | sed 's/^v//' | sed 's/_.*//')
+    END_VERSIONS
+    """
+}

modules/vsearch/usearchglobal/meta.yml

@@ -0,0 +1,38 @@
+name: "vsearch_usearchglobal"
+description: Compare target sequences to fasta-formatted query sequences using global pairwise alignment.
+keywords:
+  - vsearch
+  - usearch
+  - alignment
+  - fasta
+tools:
+  - "vsearch":
+      description: "VSEARCH is a versatile open-source tool for microbiome analysis, including chimera detection, clustering, dereplication and rereplication, extraction, FASTA/FASTQ/SFF file processing, masking, orienting, pair-wise alignment, restriction site cutting, searching, shuffling, sorting, subsampling, and taxonomic classification of amplicon sequences for metagenomics, genomics, and population genetics. (USEARCH alternative)"
+      homepage: "https://github.com/torognes/vsearch"
+      documentation: "None"
+      tool_dev_url: "https://github.com/torognes/vsearch"
+      doi: "doi: 10.7717/peerj.2584"
+      licence: "['GPL v3-or-later OR BSD-2-clause']"
+
+input:
+  - queryfasta:
+      type: file
+      description: Query sequences in FASTA format
+      pattern: "*.{fasta,fa,fna,faa}"
+  - db:
+      type: file
+      description: Reference database file. It may be in FASTA or UDB format.
+      pattern: "*"
+
+output:
+  - blast6out:
+      type: file
+      description: Tab delimited results in blast-like format
+      pattern: "*.{tsv}" 
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+
+authors:
+  - "@jtangrot"

tests/config/pytest_modules.yml

@@ -2024,6 +2024,10 @@ vcftools:
   - modules/vcftools/**
   - tests/modules/vcftools/**
 
+vsearch/usearchglobal:
+  - modules/vsearch/usearchglobal/**
+  - tests/modules/vsearch/usearchglobal/**
+
 yara/index:
   - modules/yara/index/**
   - tests/modules/yara/index/**

tests/modules/vsearch/usearchglobal/main.nf

@@ -0,0 +1,12 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { VSEARCH_USEARCHGLOBAL } from '../../../../modules/vsearch/usearchglobal/main.nf'
+
+workflow test_vsearch_usearchglobal {
+    
+    query = file(params.test_data['sarscov2']['genome']['transcriptome_fasta'], checkIfExists: true)
+    db = file(params.test_data['sarscov2']['genome']['genome_fasta'], checkIfExists: true)
+    VSEARCH_USEARCHGLOBAL ( query, db, "blast6out_results" )
+}

tests/modules/vsearch/usearchglobal/nextflow.config

@@ -0,0 +1,5 @@
+process {
+    ext.args = '--id 0.985'
+    publishDir = { "${params.outdir}/${task.process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()}" }
+    
+}

tests/modules/vsearch/usearchglobal/test.yml

@@ -0,0 +1,7 @@
+- name: vsearch usearchglobal test_vsearch_usearchglobal
+  command: nextflow run ./tests/modules/vsearch/usearchglobal -entry test_vsearch_usearchglobal -c ./tests/config/nextflow.config  -c ./tests/modules/vsearch/usearchglobal/nextflow.config
+  tags:
+    - vsearch/usearchglobal
+  files:
+    - path: output/vsearch/blast6out_results.tsv
+      md5sum: 09733131643f1d951321a6e17a35eb8c

tests/modules/vsearch/usearchglobal/test.yml.ori

@@ -0,0 +1,14 @@
+## TODO nf-core: Please run the following command to build this file:
+#                nf-core modules create-test-yml vsearch/usearchglobal
+- name: "vsearch usearchglobal"
+  command: nextflow run ./tests/modules/vsearch/usearchglobal -entry test_vsearch_usearchglobal -c ./tests/config/nextflow.config -c ./tests/modules/vsearch/usearchglobal/nextflow.config
+  tags:
+    - "vsearch"
+    #
+    - "vsearch/usearchglobal"
+    #
+  files:
+    - path: "output/vsearch/blast6out_results.tsv"
+      md5sum: e667c7caad0bc4b7ac383fd023c654fc
+    - path: output/vsearch/versions.yml
+      md5sum: a01fe51bc4c6a3a6226fbf77b2c7cf3b