Faqcs (#1339)

* Initial module creation

* Update test.yml

* Update test.yml

* Update test.yml

* Update meta.yml

Co-authored-by: Cipriano <rrn8@cdc.gov>
Co-authored-by: Maxime U. Garcia <max.u.garcia@gmail.com>

modules/faqcs/main.nf

@@ -0,0 +1,103 @@
+process FAQCS {
+    tag "$meta.id"
+    label 'process_medium'
+
+    conda (params.enable_conda ? "bioconda::faqcs=2.10" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/faqcs%3A2.10--r41h9a82719_2' :
+        'quay.io/biocontainers/faqcs:2.10--r41h9a82719_2' }"
+
+    input:
+    tuple val(meta), path(reads)
+
+    output:
+    tuple val(meta), path('*.trimmed.fastq.gz')           , emit: reads
+    tuple val(meta), path('*.stats.txt')                  , emit: stats
+    tuple val(meta), path('*_qc_report.pdf')              , optional:true, emit: statspdf
+    tuple val(meta), path('*.log')                        , emit: log
+    tuple val(meta), path('*.discard.fastq.gz')           , optional:true, emit: reads_fail
+    tuple val(meta), path('*.trimmed.unpaired.fastq.gz')  , optional:true, emit: reads_unpaired
+    path "versions.yml"                                   , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    // Added soft-links to original fastqs for consistent naming in MultiQC
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    if (meta.single_end) {
+        """
+        [ ! -f  ${prefix}.fastq.gz ] && ln -s $reads ${prefix}.fastq.gz
+        FaQCs \\
+            -d . \\
+            -u ${prefix}.fastq.gz \\
+            --prefix ${prefix} \\
+            -t $task.cpus \\
+            $args \\
+            2> ${prefix}.fastp.log
+
+
+        if [[ -f ${prefix}.unpaired.trimmed.fastq ]]; then
+            mv ${prefix}.unpaired.trimmed.fastq ${prefix}.trimmed.fastq
+            gzip ${prefix}.trimmed.fastq
+        fi
+        if [[ -f ${prefix}.discard.trimmed.fastq ]]; then
+            mv ${prefix}.discard.trimmed.fastq ${prefix}.trimmed.discard.fastq
+            gzip ${prefix}.trimmed.discard.fastq
+        fi
+        cat <<-END_VERSIONS > versions.yml
+        "${task.process}":
+            faqcs: \$(echo \$(FaQCs --version 2>&1) | sed 's/^.*Version: //;' )
+        END_VERSIONS
+        """
+    } else {
+        """
+        [ ! -f  ${prefix}_1.fastq.gz ] && ln -s ${reads[0]} ${prefix}_1.fastq.gz
+        [ ! -f  ${prefix}_2.fastq.gz ] && ln -s ${reads[1]} ${prefix}_2.fastq.gz
+        FaQCs \\
+            -d . \\
+            -1 ${prefix}_1.fastq.gz \\
+            -2 ${prefix}_2.fastq.gz \\
+            --prefix ${meta.id} \\
+            -t $task.cpus \\
+            $args \\
+            2> ${prefix}.fastp.log
+
+        # Unpaired
+        if [[ -f ${prefix}.unpaired.trimmed.fastq ]]; then
+            # If it is empty remove it
+            if [[ ! -s ${prefix}.unpaired.trimmed.fastq ]]; then
+                rm ${prefix}.unpaired.trimmed.fastq
+            else
+                mv ${prefix}.unpaired.trimmed.fastq ${prefix}.trimmed.unpaired.fastq
+                gzip ${prefix}.trimmed.unpaired.fastq
+            fi
+        fi
+
+        # R1
+        if [[ -f ${prefix}.1.trimmed.fastq ]]; then
+            mv ${prefix}.1.trimmed.fastq ${prefix}_1.trimmed.fastq
+            gzip ${prefix}_1.trimmed.fastq
+        fi
+
+        # R2
+        if [[ -f ${prefix}.2.trimmed.fastq ]]; then
+            mv ${prefix}.2.trimmed.fastq ${prefix}_2.trimmed.fastq
+            gzip ${prefix}_2.trimmed.fastq
+        fi
+
+        # Discarded: Created if --discard argument is passed
+        if [[ -f ${prefix}.discard.trimmed.fastq ]]; then
+            mv ${prefix}.discard.trimmed.fastq ${prefix}.trimmed.discard.fastq
+            gzip ${prefix}.trimmed.discard.fastq
+        fi
+
+        cat <<-END_VERSIONS > versions.yml
+        "${task.process}":
+            faqcs: \$(echo \$(FaQCs --version 2>&1) | sed 's/^.*Version: //;' )
+        END_VERSIONS
+        """
+    }
+}
+

modules/faqcs/meta.yml

@@ -0,0 +1,68 @@
+name: faqcs
+description: Perform adapter and quality trimming on sequencing reads with reporting
+keywords:
+  - trimming
+  - quality control
+  - fastq
+  - faqcs
+tools:
+  - faqcs:
+      description: |
+        FaQCs combines several features of currently available applications into a single, user-friendly process, and includes additional unique capabilities such as filtering the PhiX control sequences, conversion of FASTQ formats, and multi-threading. The original data and trimmed summaries are reported within a variety of graphics and reports, providing a simple way to do data quality control and assurance.
+      homepage: https://github.com/LANL-Bioinformatics/FaQCs
+      documentation: https://github.com/LANL-Bioinformatics/FaQCs
+      tool_dev_url: https://github.com/LANL-Bioinformatics/FaQCs
+      doi: "https://doi.org/10.1186/s12859-014-0366-2"
+      licence: ['GPLv3 License']
+
+## 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 ]
+  - reads:
+      type: file
+      description: |
+        List of input FastQ files of size 1 and 2 for single-end and paired-end data,
+        respectively.
+
+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"
+  - reads:
+      type: file
+      description: The trimmed/modified fastq reads
+      pattern: "*trimmed.fastq.gz"
+  - reads_fail:
+      type: file
+      description: Reads that failed the preprocessing (Optional with --discard args setting)
+      pattern: "*discard.fastq.gz"
+  - reads_unpaired:
+      type: file
+      description: Reads without matching mates in paired-end files (Optional)
+      pattern: "*trimmed.unpaired.fastq.gz"
+  - stats:
+      type: file
+      description: trimming/qc text stats file
+      pattern: "*.stats.txt"
+  - statspdf:
+      type: file
+      description: trimming/qc pdf report file
+      pattern: "*_qc_report.pdf"
+  - log:
+      type: file
+      description: fastq log file
+      pattern: "*.log"
+authors:
+  - "@mjcipriano"
+  - "@sateeshperi"
+  - "@hseabolt"

tests/config/pytest_modules.yml

@@ -504,6 +504,10 @@ expansionhunter:
   - modules/expansionhunter/**
   - tests/modules/expansionhunter/**
 
+faqcs:
+  - modules/faqcs/**
+  - tests/modules/faqcs/**
+
 fargene:
   - modules/fargene/**
   - tests/modules/fargene/**

tests/modules/faqcs/main.nf

@@ -0,0 +1,30 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { FAQCS } from '../../../modules/faqcs/main.nf'
+
+
+//
+// Test with single-end data
+//
+workflow test_fastp_single_end {
+    input = [ [ id:'test', single_end:true ], // meta map
+              [ file(params.test_data['sarscov2']['illumina']['test_1_fastq_gz'], checkIfExists: true) ]
+            ]
+
+    FAQCS ( input )
+}
+
+//
+// Test with paired-end data
+//
+workflow test_fastp_paired_end {
+    input = [ [ id:'test', single_end:false ], // meta map
+              [ file(params.test_data['sarscov2']['illumina']['test_1_fastq_gz'], checkIfExists: true),
+                file(params.test_data['sarscov2']['illumina']['test_2_fastq_gz'], checkIfExists: true) ]
+            ]
+
+    FAQCS ( input )
+}
+

tests/modules/faqcs/nextflow.config

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

tests/modules/faqcs/test.yml

@@ -0,0 +1,31 @@
+- name: faqcs test_fastp_single_end
+  command: nextflow run tests/modules/faqcs -entry test_fastp_single_end -c tests/config/nextflow.config
+  tags:
+    - faqcs
+  files:
+    - path: output/faqcs/test.fastp.log
+      md5sum: be79dc893f87de1f82faf749cdfb848c
+    - path: output/faqcs/test.stats.txt
+      md5sum: ea20e93706b2e4c676004253baa3cec6
+    - path: output/faqcs/test.trimmed.fastq.gz
+      md5sum: 875863b402f67403dac63ef59b9c9a8a
+    - path: output/faqcs/test_qc_report.pdf
+    - path: output/faqcs/versions.yml
+      md5sum: 2a38d7e7ab5299336e9669c393c9da6c
+
+- name: faqcs test_fastp_paired_end
+  command: nextflow run tests/modules/faqcs -entry test_fastp_paired_end -c tests/config/nextflow.config
+  tags:
+    - faqcs
+  files:
+    - path: output/faqcs/test.fastp.log
+      md5sum: be79dc893f87de1f82faf749cdfb848c
+    - path: output/faqcs/test.stats.txt
+      md5sum: 9a693f8af94ab8c485519d9a523aa622
+    - path: output/faqcs/test_1.trimmed.fastq.gz
+      md5sum: 875863b402f67403dac63ef59b9c9a8a
+    - path: output/faqcs/test_2.trimmed.fastq.gz
+      md5sum: 375aeb74819ca3d72203135ac80df78c
+    - path: output/faqcs/test_qc_report.pdf
+    - path: output/faqcs/versions.yml
+      md5sum: 208d54c0cf6dfc54e719b81b990afac9