Merge remote-tracking branch 'nf-core/master'

modules/bamtools/split/main.nf

@@ -35,7 +35,7 @@ process BAMTOOLS_SPLIT {
 
     cat <<-END_VERSIONS > versions.yml
     ${getProcessName(task.process)}:
-        bamtools: \$( bamtools --version | grep -e 'bamtools' | sed 's/^.*bamtools //' )
+        ${getSoftwareName(task.process)}: \$( bamtools --version | grep -e 'bamtools' | sed 's/^.*bamtools //' )
     END_VERSIONS
     """
 }

modules/bandage/image/meta.yml

@@ -11,6 +11,7 @@ tools:
             Bandage - a Bioinformatics Application for Navigating De novo Assembly Graphs Easily
         homepage: https://github.com/rrwick/Bandage
         documentation: https://github.com/rrwick/Bandage
+        licence: ['GPL-3.0-or-later']
 input:
     - meta:
         type: map

modules/bcftools/concat/meta.yml

@@ -13,6 +13,7 @@ tools:
         homepage: http://samtools.github.io/bcftools/bcftools.html
         documentation: http://www.htslib.org/doc/bcftools.html
         doi: 10.1093/bioinformatics/btp352
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bcftools/consensus/meta.yml

@@ -11,6 +11,7 @@ tools:
         homepage: http://samtools.github.io/bcftools/bcftools.html
         documentation: http://www.htslib.org/doc/bcftools.html
         doi: 10.1093/bioinformatics/btp352
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bcftools/filter/meta.yml

@@ -11,6 +11,7 @@ tools:
         homepage: http://samtools.github.io/bcftools/bcftools.html
         documentation: http://www.htslib.org/doc/bcftools.html
         doi: 10.1093/bioinformatics/btp352
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bcftools/index/meta.yml

@@ -13,7 +13,7 @@ tools:
       documentation: https://samtools.github.io/bcftools/howtos/index.html
       tool_dev_url: https://github.com/samtools/bcftools
       doi: "10.1093/gigascience/giab008"
-      licence: ['GPL']
+      licence: ['MIT', 'GPL-3.0-or-later']
 
 input:
   - meta:

modules/bcftools/isec/meta.yml

@@ -13,6 +13,7 @@ tools:
         homepage: http://samtools.github.io/bcftools/bcftools.html
         documentation: http://www.htslib.org/doc/bcftools.html
         doi: 10.1093/bioinformatics/btp352
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bcftools/merge/meta.yml

@@ -11,6 +11,7 @@ tools:
         homepage: http://samtools.github.io/bcftools/bcftools.html
         documentation: http://www.htslib.org/doc/bcftools.html
         doi: 10.1093/bioinformatics/btp352
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bcftools/mpileup/meta.yml

@@ -11,6 +11,7 @@ tools:
         homepage: http://samtools.github.io/bcftools/bcftools.html
         documentation: http://www.htslib.org/doc/bcftools.html
         doi: 10.1093/bioinformatics/btp352
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bcftools/norm/meta.yml

@@ -12,6 +12,7 @@ tools:
         homepage: http://samtools.github.io/bcftools/bcftools.html
         documentation: http://www.htslib.org/doc/bcftools.html
         doi: 10.1093/bioinformatics/btp352
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bcftools/query/meta.yml

@@ -12,6 +12,7 @@ tools:
         homepage: http://samtools.github.io/bcftools/bcftools.html
         documentation: http://www.htslib.org/doc/bcftools.html
         doi: 10.1093/bioinformatics/btp352
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bcftools/reheader/meta.yml

@@ -11,7 +11,7 @@ tools:
       homepage: http://samtools.github.io/bcftools/bcftools.html
       documentation: http://samtools.github.io/bcftools/bcftools.html#reheader
       doi: 10.1093/gigascience/giab008
-      licence: ['GPL']
+      licence: ['MIT']
 
 input:
   - meta:

modules/bcftools/stats/meta.yml

@@ -12,6 +12,7 @@ tools:
         homepage: http://samtools.github.io/bcftools/bcftools.html
         documentation: http://www.htslib.org/doc/bcftools.html
         doi: 10.1093/bioinformatics/btp352
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bcftools/view/meta.yml

@@ -13,6 +13,7 @@ tools:
         homepage: http://samtools.github.io/bcftools/bcftools.html
         documentation: http://www.htslib.org/doc/bcftools.html
         doi: 10.1093/bioinformatics/btp352
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bedtools/bamtobed/meta.yml

@@ -8,6 +8,7 @@ tools:
         description: |
             A set of tools for genomic analysis tasks, specifically enabling genome arithmetic (merge, count, complement) on various file types.
         documentation: https://bedtools.readthedocs.io/en/latest/content/tools/complement.html
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bedtools/complement/meta.yml

@@ -8,6 +8,7 @@ tools:
         description: |
             A set of tools for genomic analysis tasks, specifically enabling genome arithmetic (merge, count, complement) on various file types.
         documentation: https://bedtools.readthedocs.io/en/latest/content/tools/complement.html
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bedtools/genomecov/meta.yml

@@ -9,6 +9,7 @@ tools:
         description: |
             A set of tools for genomic analysis tasks, specifically enabling genome arithmetic (merge, count, complement) on various file types.
         documentation: https://bedtools.readthedocs.io/en/latest/content/tools/genomecov.html
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bedtools/getfasta/meta.yml

@@ -9,6 +9,7 @@ tools:
         description: |
             A set of tools for genomic analysis tasks, specifically enabling genome arithmetic (merge, count, complement) on various file types.
         documentation: https://bedtools.readthedocs.io/en/latest/content/tools/intersect.html
+        licence: ['MIT']
 input:
     - bed:
         type: file

modules/bedtools/intersect/meta.yml

@@ -8,6 +8,7 @@ tools:
         description: |
             A set of tools for genomic analysis tasks, specifically enabling genome arithmetic (merge, count, complement) on various file types.
         documentation: https://bedtools.readthedocs.io/en/latest/content/tools/intersect.html
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bedtools/makewindows/meta.yml

@@ -11,7 +11,7 @@ tools:
       documentation: https://bedtools.readthedocs.io/en/latest/content/tools/makewindows.html
       tool_dev_url: None
       doi: "10.1093/bioinformatics/btq033"
-      licence: ['GPL v2']
+      licence: ['MIT']
 input:
   - meta:
       type: map

modules/bedtools/maskfasta/meta.yml

@@ -9,6 +9,7 @@ tools:
         description: |
             A set of tools for genomic analysis tasks, specifically enabling genome arithmetic (merge, count, complement) on various file types.
         documentation: https://bedtools.readthedocs.io/en/latest/content/tools/intersect.html
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bedtools/merge/meta.yml

@@ -8,6 +8,7 @@ tools:
         description: |
             A set of tools for genomic analysis tasks, specifically enabling genome arithmetic (merge, count, complement) on various file types.
         documentation: https://bedtools.readthedocs.io/en/latest/content/tools/merge.html
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bedtools/slop/meta.yml

@@ -8,6 +8,7 @@ tools:
         description: |
             A set of tools for genomic analysis tasks, specifically enabling genome arithmetic (merge, count, complement) on various file types.
         documentation: https://bedtools.readthedocs.io/en/latest/content/tools/slop.html
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bedtools/sort/meta.yml

@@ -8,6 +8,7 @@ tools:
         description: |
             A set of tools for genomic analysis tasks, specifically enabling genome arithmetic (merge, count, complement) on various file types.
         documentation: https://bedtools.readthedocs.io/en/latest/content/tools/sort.html
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bedtools/subtract/meta.yml

@@ -10,6 +10,7 @@ tools:
         description: |
             A set of tools for genomic analysis tasks, specifically enabling genome arithmetic (merge, count, complement) on various file types.
         documentation: https://bedtools.readthedocs.io/en/latest/content/tools/subtract.html
+        licence: ['MIT']
 
 input:
   - meta:

modules/bismark/align/meta.yml

@@ -17,6 +17,7 @@ tools:
         homepage: https://github.com/FelixKrueger/Bismark
         documentation: https://github.com/FelixKrueger/Bismark/tree/master/Docs
         doi: 10.1093/bioinformatics/btr167
+        licence: ['GPL-3.0-or-later']
 input:
     - meta:
         type: map

modules/bismark/deduplicate/meta.yml

@@ -19,6 +19,7 @@ tools:
         homepage: https://github.com/FelixKrueger/Bismark
         documentation: https://github.com/FelixKrueger/Bismark/tree/master/Docs
         doi: 10.1093/bioinformatics/btr167
+        licence: ['GPL-3.0-or-later']
 input:
     - meta:
         type: map

modules/bismark/genomepreparation/meta.yml

@@ -19,6 +19,7 @@ tools:
         homepage: https://github.com/FelixKrueger/Bismark
         documentation: https://github.com/FelixKrueger/Bismark/tree/master/Docs
         doi: 10.1093/bioinformatics/btr167
+        licence: ['GPL-3.0-or-later']
 input:
     - fasta:
         type: file

modules/bismark/methylationextractor/meta.yml

@@ -18,6 +18,7 @@ tools:
         homepage: https://github.com/FelixKrueger/Bismark
         documentation: https://github.com/FelixKrueger/Bismark/tree/master/Docs
         doi: 10.1093/bioinformatics/btr167
+        licence: ['GPL-3.0-or-later']
 input:
     - meta:
         type: map

modules/bismark/report/meta.yml

@@ -16,6 +16,7 @@ tools:
         homepage: https://github.com/FelixKrueger/Bismark
         documentation: https://github.com/FelixKrueger/Bismark/tree/master/Docs
         doi: 10.1093/bioinformatics/btr167
+        licence: ['GPL-3.0-or-later']
 input:
     - meta:
         type: map

modules/bismark/summary/meta.yml

@@ -19,6 +19,7 @@ tools:
         homepage: https://github.com/FelixKrueger/Bismark
         documentation: https://github.com/FelixKrueger/Bismark/tree/master/Docs
         doi: 10.1093/bioinformatics/btr167
+        licence: ['GPL-3.0-or-later']
 input:
     - bam:
         type: file

modules/blast/blastn/meta.yml

@@ -12,6 +12,7 @@ tools:
         homepage: https://blast.ncbi.nlm.nih.gov/Blast.cgi
         documentation: https://blast.ncbi.nlm.nih.gov/Blast.cgi?CMD=Web&PAGE_TYPE=Blastdocs
         doi: 10.1016/S0022-2836(05)80360-2
+        licence: ['US-Government-Work']
 input:
     - meta:
         type: map

modules/blast/makeblastdb/meta.yml

@@ -11,6 +11,7 @@ tools:
           homepage: https://blast.ncbi.nlm.nih.gov/Blast.cgi
           documentation: https://blast.ncbi.nlm.nih.gov/Blast.cgi?CMD=Web&PAGE_TYPE=Blastdocs
           doi: 10.1016/S0022-2836(05)80360-2
+          licence: ['US-Government-Work']
 input:
     - fasta:
         type: file

modules/bowtie/align/meta.yml

@@ -13,6 +13,7 @@ tools:
           homepage: http://bowtie-bio.sourceforge.net/index.shtml
           documentation: http://bowtie-bio.sourceforge.net/manual.shtml
           arxiv: arXiv:1303.3997
+          licence: ['Artistic-2.0']
 input:
     - meta:
           type: map

modules/bowtie/build/meta.yml

@@ -13,6 +13,7 @@ tools:
           homepage: http://bowtie-bio.sourceforge.net/index.shtml
           documentation: http://bowtie-bio.sourceforge.net/manual.shtml
           arxiv: arXiv:1303.3997
+          licence: ['Artistic-2.0']
 input:
     - fasta:
           type: file

modules/bowtie2/align/meta.yml

@@ -13,6 +13,7 @@ tools:
           homepage: http://bowtie-bio.sourceforge.net/bowtie2/index.shtml
           documentation: http://bowtie-bio.sourceforge.net/bowtie2/manual.shtml
           doi: 10.1038/nmeth.1923
+          licence: ['GPL-3.0-or-later']
 input:
     - meta:
           type: map

modules/bowtie2/build/meta.yml

@@ -14,6 +14,7 @@ tools:
           homepage: http://bowtie-bio.sourceforge.net/bowtie2/index.shtml
           documentation: http://bowtie-bio.sourceforge.net/bowtie2/manual.shtml
           doi: 10.1038/nmeth.1923
+          licence: ['GPL-3.0-or-later']
 input:
     - fasta:
           type: file

modules/bwa/aln/meta.yml

@@ -17,7 +17,7 @@ tools:
         homepage: http://bio-bwa.sourceforge.net/
         documentation: http://bio-bwa.sourceforge.net/
         doi: "10.1093/bioinformatics/btp324"
-      licence: ['GPL v3']
+        licence: ['GPL-3.0-or-later']
 
 input:
     - meta:

modules/bwa/index/meta.yml

@@ -13,6 +13,7 @@ tools:
         homepage: http://bio-bwa.sourceforge.net/
         documentation: http://www.htslib.org/doc/samtools.html
         arxiv: arXiv:1303.3997
+        licence: ['GPL-3.0-or-later']
 input:
     - fasta:
         type: file

modules/bwa/mem/meta.yml

@@ -16,6 +16,7 @@ tools:
         homepage: http://bio-bwa.sourceforge.net/
         documentation: http://www.htslib.org/doc/samtools.html
         arxiv: arXiv:1303.3997
+        licence: ['GPL-3.0-or-later']
 input:
     - meta:
         type: map

modules/bwa/sampe/meta.yml

@@ -18,7 +18,7 @@ tools:
         homepage: http://bio-bwa.sourceforge.net/
         documentation: http://bio-bwa.sourceforge.net/
         doi: "10.1093/bioinformatics/btp324"
-      licence: ['GPL v3']
+        licence: ['GPL-3.0-or-later']
 
 input:
   - meta:

modules/bwa/samse/meta.yml

@@ -19,7 +19,7 @@ tools:
         homepage: http://bio-bwa.sourceforge.net/
         documentation: http://bio-bwa.sourceforge.net/
         doi: "10.1093/bioinformatics/btp324"
-      licence: ['GPL v3']
+        licence: ['GPL-3.0-or-later']
 
 input:
   - meta:

modules/bwamem2/index/meta.yml

@@ -12,6 +12,7 @@ tools:
             a large reference genome, such as the human genome.
         homepage: https://github.com/bwa-mem2/bwa-mem2
         documentation: https://github.com/bwa-mem2/bwa-mem2#usage
+        licence: ['MIT']
 input:
     - fasta:
         type: file

modules/bwamem2/mem/meta.yml

@@ -16,6 +16,7 @@ tools:
         homepage: http://bio-bwa.sourceforge.net/
         documentation: http://www.htslib.org/doc/samtools.html
         arxiv: arXiv:1303.3997
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bwameth/align/meta.yml

@@ -19,6 +19,7 @@ tools:
         homepage: https://github.com/brentp/bwa-meth
         documentation: https://github.com/brentp/bwa-meth
         arxiv: arXiv:1401.1129
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/bwameth/index/meta.yml

@@ -15,6 +15,7 @@ tools:
         homepage: https://github.com/brentp/bwa-meth
         documentation: https://github.com/brentp/bwa-meth
         arxiv: arXiv:1401.1129
+        licence: ['MIT']
 input:
     - fasta:
         type: file

modules/cat/cat/meta.yml

@@ -10,7 +10,7 @@ tools:
       homepage: None
       documentation: https://man7.org/linux/man-pages/man1/cat.1.html
       tool_dev_url: None
-
+      licence: ['GPL-3.0-or-later']
 input:
   - files_in:
       type: file

modules/cat/fastq/meta.yml

@@ -8,6 +8,7 @@ tools:
         description: |
             The cat utility reads files sequentially, writing them to the standard output.
         documentation: https://www.gnu.org/software/coreutils/manual/html_node/cat-invocation.html
+        licence: ['GPL-3.0-or-later']
 input:
     - meta:
         type: map

modules/checkm/lineagewf/functions.nf

@@ -0,0 +1,78 @@
+//
+//  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()
+}
+
+//
+// Extract name of module from process name using $task.process
+//
+def getProcessName(task_process) {
+    return task_process.tokenize(':')[-1]
+}
+
+//
+// Function to initialise default values and to generate a Groovy Map of available options for nf-core modules
+//
+def initOptions(Map args) {
+    def Map options = [:]
+    options.args            = args.args ?: ''
+    options.args2           = args.args2 ?: ''
+    options.args3           = args.args3 ?: ''
+    options.publish_by_meta = args.publish_by_meta ?: []
+    options.publish_dir     = args.publish_dir ?: ''
+    options.publish_files   = args.publish_files
+    options.suffix          = args.suffix ?: ''
+    return options
+}
+
+//
+// Tidy up and join elements of a list to return a path string
+//
+def getPathFromList(path_list) {
+    def paths = path_list.findAll { item -> !item?.trim().isEmpty() }      // Remove empty entries
+    paths     = paths.collect { it.trim().replaceAll("^[/]+|[/]+\$", "") } // Trim whitespace and trailing slashes
+    return paths.join('/')
+}
+
+//
+// Function to save/publish module results
+//
+def saveFiles(Map args) {
+    def ioptions  = initOptions(args.options)
+    def path_list = [ ioptions.publish_dir ?: args.publish_dir ]
+
+    // Do not publish versions.yml unless running from pytest workflow
+    if (args.filename.equals('versions.yml') && !System.getenv("NF_CORE_MODULES_TEST")) {
+        return null
+    }
+    if (ioptions.publish_by_meta) {
+        def key_list = ioptions.publish_by_meta instanceof List ? ioptions.publish_by_meta : args.publish_by_meta
+        for (key in key_list) {
+            if (args.meta && key instanceof String) {
+                def path = key
+                if (args.meta.containsKey(key)) {
+                    path = args.meta[key] instanceof Boolean ? "${key}_${args.meta[key]}".toString() : args.meta[key]
+                }
+                path = path instanceof String ? path : ''
+                path_list.add(path)
+            }
+        }
+    }
+    if (ioptions.publish_files instanceof Map) {
+        for (ext in ioptions.publish_files) {
+            if (args.filename.endsWith(ext.key)) {
+                def ext_list = path_list.collect()
+                ext_list.add(ext.value)
+                return "${getPathFromList(ext_list)}/$args.filename"
+            }
+        }
+    } else if (ioptions.publish_files == null) {
+        return "${getPathFromList(path_list)}/$args.filename"
+    }
+}

modules/checkm/lineagewf/main.nf

@@ -0,0 +1,49 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName; getProcessName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process CHECKM_LINEAGEWF {
+    tag "$meta.id"
+    label 'process_medium'
+    publishDir "${params.outdir}",
+        mode: params.publish_dir_mode,
+        saveAs: { filename -> saveFiles(filename:filename, options:params.options, publish_dir:getSoftwareName(task.process), meta:meta, publish_by_meta:['id']) }
+
+    conda (params.enable_conda ? "bioconda::checkm-genome=1.1.3" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/checkm-genome:1.1.3--py_1"
+    } else {
+        container "quay.io/biocontainers/checkm-genome:1.1.3--py_1"
+    }
+
+    input:
+    tuple val(meta), path(fasta)
+    val fasta_ext
+
+    output:
+    tuple val(meta), path("${prefix}")    , emit: checkm_output
+    tuple val(meta), path("${prefix}.tsv"), emit: checkm_tsv
+    path "versions.yml"                   , emit: versions
+
+    script:
+    prefix = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    """
+    checkm \\
+        lineage_wf \\
+        -t $task.cpus \\
+        -f ${prefix}.tsv \\
+        --tab_table \\
+        --pplacer_threads $task.cpus \\
+        -x $fasta_ext \\
+        $options.args \\
+        . \\
+        $prefix
+
+    cat <<-END_VERSIONS > versions.yml
+    ${getProcessName(task.process)}:
+        ${getSoftwareName(task.process)}: \$( checkm 2>&1 | grep '...:::' | sed 's/.*CheckM v//;s/ .*//' )
+    END_VERSIONS
+    """
+}

modules/checkm/lineagewf/meta.yml

@@ -0,0 +1,58 @@
+name: checkm_lineagewf
+description: CheckM provides a set of tools for assessing the quality of genomes recovered from isolates, single cells, or metagenomes.
+keywords:
+  - checkm
+  - mag
+  - metagenome
+  - quality
+  - isolates
+  - microbes
+  - single cells
+  - completeness
+  - contamination
+  - bins
+  - genome bins
+tools:
+  - checkm:
+      description: Assess the quality of microbial genomes recovered from isolates, single cells, and metagenomes.
+      homepage: https://ecogenomics.github.io/CheckM/
+      documentation: https://github.com/Ecogenomics/CheckM/wiki
+      tool_dev_url: https://github.com/Ecogenomics/CheckM
+      doi: "10.1101/gr.186072.114"
+      licence: ['GPL v3']
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - fasta:
+      type: file
+      description: One or a list of multiple FASTA files of each bin, with extension defined with the fasta_ext value
+      pattern: "*.{$fasta_ext}"
+  - fasta_ext:
+      type: value
+      description: The file-type extension suffix of the input FASTA files (e.g., fasta, fna, fa, fas)
+
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'sample', bin:'1' ]
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+  - checkm_output:
+      type: directory
+      description: CheckM output directory
+      pattern: "*/"
+  - checkm_tsv:
+      type: file
+      description: CheckM summary completeness statistics table
+      pattern: "*.tsv"
+
+authors:
+  - "@jfy133"

modules/cnvkit/meta.yml

@@ -10,6 +10,7 @@ tools:
         CNVkit is a Python library and command-line software toolkit to infer and visualize copy number from high-throughput DNA sequencing data. It is designed for use with hybrid capture, including both whole-exome and custom target panels, and short-read sequencing platforms such as Illumina and Ion Torrent.
       homepage: https://cnvkit.readthedocs.io/en/stable/index.html
       documentation: https://cnvkit.readthedocs.io/en/stable/index.html
+      licence: ['Apache-2.0']
 params:
   - outdir:
       type: string

modules/cooler/digest/meta.yml

@@ -10,7 +10,7 @@ tools:
       documentation: https://cooler.readthedocs.io/en/latest/index.html
       tool_dev_url: https://github.com/open2c/cooler
       doi: "10.1093/bioinformatics/btz540"
-      licence: ['BSD-3-clause']
+      licence: ['BSD-3-Clause']
 
 input:
   - fasta:

modules/cooler/dump/meta.yml

@@ -9,7 +9,7 @@ tools:
       documentation: https://cooler.readthedocs.io/en/latest/index.html
       tool_dev_url: https://github.com/open2c/cooler
       doi: "10.1093/bioinformatics/btz540"
-      licence: ['BSD-3-clause']
+      licence: ['BSD-3-Clause']
 
 input:
   - meta:

modules/custom/dumpsoftwareversions/meta.yml

@@ -8,7 +8,7 @@ tools:
       description: Custom module used to dump software versions within the nf-core pipeline template
       homepage: https://github.com/nf-core/tools
       documentation: https://github.com/nf-core/tools
-
+      licence: ['MIT']
 input:
   - versions:
       type: file

modules/cutadapt/meta.yml

@@ -11,6 +11,7 @@ tools:
         Cutadapt finds and removes adapter sequences, primers, poly-A tails and other types of unwanted sequence from your high-throughput sequencing reads.
       documentation: https://cutadapt.readthedocs.io/en/stable/index.html
       doi: DOI:10.14806/ej.17.1.200
+      licence: ['MIT']
 input:
   - meta:
       type: map

modules/delly/call/meta.yml

@@ -13,7 +13,7 @@ tools:
       documentation: https://github.com/dellytools/delly/blob/master/README.md
       tool_dev_url: None
       doi: "DOI:10.1093/bioinformatics/bts378"
-      licence: ["BSD-3-clause"]
+      licence: ['BSD-3-Clause']
 
 input:
   - meta:

modules/dshbio/exportsegments/meta.yml

@@ -12,6 +12,7 @@ tools:
             or later.
         homepage: https://github.com/heuermh/dishevelled-bio
         documentation: https://github.com/heuermh/dishevelled-bio
+        licence: ['LGPL-3.0-or-later']
 input:
     - meta:
         type: map

modules/dshbio/filterbed/meta.yml

@@ -10,6 +10,7 @@ tools:
             or later.
         homepage: https://github.com/heuermh/dishevelled-bio
         documentation: https://github.com/heuermh/dishevelled-bio
+        licence: ['LGPL-3.0-or-later']
 input:
     - meta:
         type: map

modules/dshbio/filtergff3/meta.yml

@@ -10,6 +10,7 @@ tools:
             or later.
         homepage: https://github.com/heuermh/dishevelled-bio
         documentation: https://github.com/heuermh/dishevelled-bio
+        licence: ['LGPL-3.0-or-later']
 input:
     - meta:
         type: map

modules/dshbio/splitbed/meta.yml

@@ -10,6 +10,7 @@ tools:
             or later.
         homepage: https://github.com/heuermh/dishevelled-bio
         documentation: https://github.com/heuermh/dishevelled-bio
+        licence: ['LGPL-3.0-or-later']
 input:
     - meta:
         type: map

modules/dshbio/splitgff3/meta.yml

@@ -10,6 +10,7 @@ tools:
             or later.
         homepage: https://github.com/heuermh/dishevelled-bio
         documentation: https://github.com/heuermh/dishevelled-bio
+        licence: ['LGPL-3.0-or-later']
 input:
     - meta:
         type: map

modules/ensemblvep/meta.yml

@@ -9,6 +9,7 @@ tools:
             or structural variants) on genes, transcripts, and protein sequence, as well as regulatory regions.
         homepage: https://www.ensembl.org/info/docs/tools/vep/index.html
         documentation: https://www.ensembl.org/info/docs/tools/vep/script/index.html
+        licence: ['Apache-2.0']
 params:
     - use_cache:
         type: boolean

modules/expansionhunter/meta.yml

@@ -10,7 +10,7 @@ tools:
       documentation: https://github.com/Illumina/ExpansionHunter/blob/master/docs/01_Introduction.md
       tool_dev_url: None
       doi: "10.1093/bioinformatics/btz431"
-      licence: ['Apache v2.0']
+      licence: ['Apache-2.0']
 
 input:
   - meta:

modules/fastp/meta.yml

@@ -10,6 +10,7 @@ tools:
         A tool designed to provide fast all-in-one preprocessing for FastQ files. This tool is developed in C++ with multithreading supported to afford high performance.
       documentation: https://github.com/OpenGene/fastp
       doi: https://doi.org/10.1093/bioinformatics/bty560
+      licence: ['MIT']
 input:
   - meta:
       type: map

modules/fastqc/meta.yml

@@ -15,6 +15,7 @@ tools:
             overrepresented sequences.
         homepage: https://www.bioinformatics.babraham.ac.uk/projects/fastqc/
         documentation: https://www.bioinformatics.babraham.ac.uk/projects/fastqc/Help/
+        licence: ['GPL-2.0-only']
 input:
     - meta:
         type: map

modules/gatk4/applybqsr/meta.yml

@@ -12,6 +12,7 @@ tools:
     homepage: https://gatk.broadinstitute.org/hc/en-us
     documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
     doi: 10.1158/1538-7445.AM2017-3590
+    licence: ['Apache-2.0']
 
 input:
   - meta:

modules/gatk4/baserecalibrator/meta.yml

@@ -11,6 +11,7 @@ tools:
     homepage: https://gatk.broadinstitute.org/hc/en-us
     documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
     doi: 10.1158/1538-7445.AM2017-3590
+    licence: ['Apache-2.0']
 
 
 input:

modules/gatk4/bedtointervallist/meta.yml

@@ -12,6 +12,7 @@ tools:
       homepage: https://gatk.broadinstitute.org/hc/en-us
       documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
       doi: 10.1158/1538-7445.AM2017-3590
+      licence: ['Apache-2.0']
 input:
   - meta:
       type: map

modules/gatk4/calculatecontamination/meta.yml

@@ -16,6 +16,7 @@ tools:
       homepage: https://gatk.broadinstitute.org/hc/en-us
       documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
       doi: 10.1158/1538-7445.AM2017-3590
+      licence: ['Apache-2.0']
 
 input:
   - meta:

modules/gatk4/createsequencedictionary/meta.yml

@@ -12,6 +12,8 @@ tools:
         homepage: https://gatk.broadinstitute.org/hc/en-us
         documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
         doi: 10.1158/1538-7445.AM2017-3590
+        licence: ['Apache-2.0']
+
 input:
     - fasta:
         type: file

modules/gatk4/createsomaticpanelofnormals/functions.nf

@@ -0,0 +1,78 @@
+//
+//  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()
+}
+
+//
+// Extract name of module from process name using $task.process
+//
+def getProcessName(task_process) {
+    return task_process.tokenize(':')[-1]
+}
+
+//
+// Function to initialise default values and to generate a Groovy Map of available options for nf-core modules
+//
+def initOptions(Map args) {
+    def Map options = [:]
+    options.args            = args.args ?: ''
+    options.args2           = args.args2 ?: ''
+    options.args3           = args.args3 ?: ''
+    options.publish_by_meta = args.publish_by_meta ?: []
+    options.publish_dir     = args.publish_dir ?: ''
+    options.publish_files   = args.publish_files
+    options.suffix          = args.suffix ?: ''
+    return options
+}
+
+//
+// Tidy up and join elements of a list to return a path string
+//
+def getPathFromList(path_list) {
+    def paths = path_list.findAll { item -> !item?.trim().isEmpty() }      // Remove empty entries
+    paths     = paths.collect { it.trim().replaceAll("^[/]+|[/]+\$", "") } // Trim whitespace and trailing slashes
+    return paths.join('/')
+}
+
+//
+// Function to save/publish module results
+//
+def saveFiles(Map args) {
+    def ioptions  = initOptions(args.options)
+    def path_list = [ ioptions.publish_dir ?: args.publish_dir ]
+
+    // Do not publish versions.yml unless running from pytest workflow
+    if (args.filename.equals('versions.yml') && !System.getenv("NF_CORE_MODULES_TEST")) {
+        return null
+    }
+    if (ioptions.publish_by_meta) {
+        def key_list = ioptions.publish_by_meta instanceof List ? ioptions.publish_by_meta : args.publish_by_meta
+        for (key in key_list) {
+            if (args.meta && key instanceof String) {
+                def path = key
+                if (args.meta.containsKey(key)) {
+                    path = args.meta[key] instanceof Boolean ? "${key}_${args.meta[key]}".toString() : args.meta[key]
+                }
+                path = path instanceof String ? path : ''
+                path_list.add(path)
+            }
+        }
+    }
+    if (ioptions.publish_files instanceof Map) {
+        for (ext in ioptions.publish_files) {
+            if (args.filename.endsWith(ext.key)) {
+                def ext_list = path_list.collect()
+                ext_list.add(ext.value)
+                return "${getPathFromList(ext_list)}/$args.filename"
+            }
+        }
+    } else if (ioptions.publish_files == null) {
+        return "${getPathFromList(path_list)}/$args.filename"
+    }
+}

modules/gatk4/createsomaticpanelofnormals/main.nf

@@ -0,0 +1,47 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName; getProcessName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process GATK4_CREATESOMATICPANELOFNORMALS {
+    tag "$meta.id"
+    label 'process_low'
+    publishDir "${params.outdir}",
+        mode: params.publish_dir_mode,
+        saveAs: { filename -> saveFiles(filename:filename, options:params.options, publish_dir:getSoftwareName(task.process), meta:meta, publish_by_meta:['id']) }
+
+    conda (params.enable_conda ? "bioconda::gatk4=4.2.0.0" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/gatk4:4.2.0.0--0"
+    } else {
+        container "quay.io/biocontainers/gatk4:4.2.0.0--0"
+    }
+
+    input:
+    tuple val(meta), path(genomicsdb)
+    path fasta
+    path fastaidx
+    path dict
+
+    output:
+    tuple val(meta), path("*.vcf.gz"), emit: vcf
+    tuple val(meta), path("*.tbi")   , emit: tbi
+    path "versions.yml"              , emit: versions
+
+    script:
+    def prefix = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    """
+    gatk \\
+        CreateSomaticPanelOfNormals \\
+        -R $fasta \\
+        -V gendb://$genomicsdb \\
+        -O ${prefix}.vcf.gz \\
+        $options.args
+
+    cat <<-END_VERSIONS > versions.yml
+    ${getProcessName(task.process)}:
+        ${getSoftwareName(task.process)}: \$(echo \$(gatk --version 2>&1) | sed 's/^.*(GATK) v//; s/ .*\$//')
+    END_VERSIONS
+    """
+}

modules/gatk4/createsomaticpanelofnormals/meta.yml

@@ -0,0 +1,55 @@
+name: gatk4_createsomaticpanelofnormals
+description: Create a panel of normals contraining germline and artifactual sites for use with mutect2.
+keywords:
+  - gatk4
+  - createsomaticpanelofnormals
+  - panelofnormals
+tools:
+  - gatk4:
+      description: |
+        Developed in the Data Sciences Platform at the Broad Institute, the toolkit offers a wide variety of tools
+        with a primary focus on variant discovery and genotyping. Its powerful processing engine
+        and high-performance computing features make it capable of taking on projects of any size.
+      homepage: https://gatk.broadinstitute.org/hc/en-us
+      documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
+      doi: 10.1158/1538-7445.AM2017-3590
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test']
+  - genoomicsdb:
+      type: directory
+      description: genomicsDB workspace that contains the samples to create the somatic panel of normals with.
+      pattern: "*_genomicsDBworkspace"
+  - fasta:
+      type: file
+      description: The reference fasta file
+      pattern: "*.fasta"
+  - fastaidx:
+      type: file
+      description: Index of reference fasta file
+      pattern: "fasta.fai"
+  - dict:
+      type: file
+      description: GATK sequence dictionary
+      pattern: "*.dict"
+
+output:
+  - vcf:
+      type: file
+      description: panel of normal as compressed vcf file
+      pattern: "*.vcf.gz"
+  - tbi:
+      type: file
+      description: Index of vcf file
+      pattern: "*vcf.gz.tbi"
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+
+authors:
+  - "@GCJMackenzie"

modules/gatk4/fastqtosam/meta.yml

@@ -1,5 +1,5 @@
 name: gatk4_fastqtosam
-description: Converts FastQ file to BAM format
+description: Converts FastQ file to SAM/BAM format
 keywords:
   - bam
   - fastq
@@ -14,7 +14,7 @@ tools:
       documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
       tool_dev_url: https://github.com/broadinstitute/gatk
       doi: "10.1158/1538-7445.AM2017-3590"
-      licence: ['BSD-3-clause']
+      licence: ['MIT']
 
 input:
   - meta:

modules/gatk4/filtermutectcalls/functions.nf

@@ -0,0 +1,78 @@
+//
+//  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()
+}
+
+//
+// Extract name of module from process name using $task.process
+//
+def getProcessName(task_process) {
+    return task_process.tokenize(':')[-1]
+}
+
+//
+// Function to initialise default values and to generate a Groovy Map of available options for nf-core modules
+//
+def initOptions(Map args) {
+    def Map options = [:]
+    options.args            = args.args ?: ''
+    options.args2           = args.args2 ?: ''
+    options.args3           = args.args3 ?: ''
+    options.publish_by_meta = args.publish_by_meta ?: []
+    options.publish_dir     = args.publish_dir ?: ''
+    options.publish_files   = args.publish_files
+    options.suffix          = args.suffix ?: ''
+    return options
+}
+
+//
+// Tidy up and join elements of a list to return a path string
+//
+def getPathFromList(path_list) {
+    def paths = path_list.findAll { item -> !item?.trim().isEmpty() }      // Remove empty entries
+    paths     = paths.collect { it.trim().replaceAll("^[/]+|[/]+\$", "") } // Trim whitespace and trailing slashes
+    return paths.join('/')
+}
+
+//
+// Function to save/publish module results
+//
+def saveFiles(Map args) {
+    def ioptions  = initOptions(args.options)
+    def path_list = [ ioptions.publish_dir ?: args.publish_dir ]
+
+    // Do not publish versions.yml unless running from pytest workflow
+    if (args.filename.equals('versions.yml') && !System.getenv("NF_CORE_MODULES_TEST")) {
+        return null
+    }
+    if (ioptions.publish_by_meta) {
+        def key_list = ioptions.publish_by_meta instanceof List ? ioptions.publish_by_meta : args.publish_by_meta
+        for (key in key_list) {
+            if (args.meta && key instanceof String) {
+                def path = key
+                if (args.meta.containsKey(key)) {
+                    path = args.meta[key] instanceof Boolean ? "${key}_${args.meta[key]}".toString() : args.meta[key]
+                }
+                path = path instanceof String ? path : ''
+                path_list.add(path)
+            }
+        }
+    }
+    if (ioptions.publish_files instanceof Map) {
+        for (ext in ioptions.publish_files) {
+            if (args.filename.endsWith(ext.key)) {
+                def ext_list = path_list.collect()
+                ext_list.add(ext.value)
+                return "${getPathFromList(ext_list)}/$args.filename"
+            }
+        }
+    } else if (ioptions.publish_files == null) {
+        return "${getPathFromList(path_list)}/$args.filename"
+    }
+}

modules/gatk4/filtermutectcalls/main.nf

@@ -0,0 +1,65 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName; getProcessName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process GATK4_FILTERMUTECTCALLS {
+    tag "$meta.id"
+    label 'process_low'
+    publishDir "${params.outdir}",
+        mode: params.publish_dir_mode,
+        saveAs: { filename -> saveFiles(filename:filename, options:params.options, publish_dir:getSoftwareName(task.process), meta:meta, publish_by_meta:['id']) }
+
+    conda (params.enable_conda ? "bioconda::gatk4=4.2.0.0" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/gatk4:4.2.0.0--0"
+    } else {
+        container "quay.io/biocontainers/gatk4:4.2.0.0--0"
+    }
+
+    input:
+    tuple val(meta), path(vcf), path(tbi), path(stats), path(orientationbias), path(segmentation), path(contaminationfile), val(contaminationest)
+    path fasta
+    path fastaidx
+    path dict
+
+    output:
+    tuple val(meta), path("*.vcf.gz")            , emit: vcf
+    tuple val(meta), path("*.vcf.gz.tbi")        , emit: tbi
+    tuple val(meta), path("*.filteringStats.tsv"), emit: stats
+    path "versions.yml"                          , emit: versions
+
+    script:
+    def prefix = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+
+    def orientationbias_options = ''
+    if (orientationbias) {
+        orientationbias_options = '--orientation-bias-artifact-priors ' + orientationbias.join(' --orientation-bias-artifact-priors ')
+    }
+
+    def segmentation_options = ''
+    if (segmentation) {
+        segmentation_options = '--tumor-segmentation ' + segmentation.join(' --tumor-segmentation ')
+    }
+
+    def contamination_options = contaminationest ? " --contamination-estimate ${contaminationest} " : ''
+    if (contaminationfile) {
+        contamination_options = '--contamination-table ' + contaminationfile.join(' --contamination-table ')
+    }
+    """
+    gatk FilterMutectCalls \\
+        -R $fasta \\
+        -V $vcf \\
+        $orientationbias_options \\
+        $segmentation_options \\
+        $contamination_options \\
+        -O ${prefix}.vcf.gz \\
+        $options.args
+
+    cat <<-END_VERSIONS > versions.yml
+    ${getProcessName(task.process)}:
+        ${getSoftwareName(task.process)}: \$(echo \$(gatk --version 2>&1) | sed 's/^.*(GATK) v//; s/ .*\$//')
+    END_VERSIONS
+    """
+}

modules/gatk4/filtermutectcalls/meta.yml

@@ -0,0 +1,84 @@
+name: gatk4_filtermutectcalls
+description: |
+        Filters the raw output of mutect2, can optionally use outputs of calculatecontamination and learnreadorientationmodel to improve filtering.
+keywords:
+  - filtermutectcalls
+  - mutect2
+  - gatk4
+  - filtervcf
+tools:
+  - gatk4:
+      description: |
+        Developed in the Data Sciences Platform at the Broad Institute, the toolkit offers a wide variety of tools
+        with a primary focus on variant discovery and genotyping. Its powerful processing engine
+        and high-performance computing features make it capable of taking on projects of any size.
+      homepage: https://gatk.broadinstitute.org/hc/en-us
+      documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
+      doi: 10.1158/1538-7445.AM2017-3590
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test' ]
+  - vcf:
+      type: file
+      description: compressed vcf file of mutect2calls
+      pattern: "*.vcf.gz"
+  - tbi:
+      type: file
+      description: Index of vcf file
+      pattern: "*vcf.gz.tbi"
+  - stats:
+      type: file
+      description: Stats file that pairs with output vcf file
+      pattern: "*vcf.gz.stats"
+  - orientationbias:
+      type: list
+      description: files containing artifact priors for input vcf. Optional input.
+      pattern: "*.artifact-prior.tar.gz"
+  - segmentation:
+      type: list
+      description: tables containing segmentation information for input vcf. Optional input.
+      pattern: "*.segmentation.table"
+  - contaminationfile:
+      type: list
+      description: table(s) containing contamination contamination data for input vcf. Optional input, takes priority over contaminationest.
+      pattern: "*.contamination.table"
+  - contaminationest:
+      type: val
+      description: estimation of contamination value as a double. Optional input, will only be used if contaminationfile is not specified.
+  - fasta:
+      type: file
+      description: The reference fasta file
+      pattern: "*.fasta"
+  - fastaidx:
+      type: file
+      description: Index of reference fasta file
+      pattern: "fasta.fai"
+  - dict:
+      type: file
+      description: GATK sequence dictionary
+      pattern: "*.dict"
+
+output:
+  - vcf:
+      type: file
+      description: file containing filtered mutect2 calls.
+      pattern: "*.vcf.gz"
+  - tbi:
+      type: file
+      description: tbi file that pairs with vcf.
+      pattern: "*.vcf.gz.tbi"
+  - stats:
+      type: file
+      description: file containing statistics of the filtermutectcalls run.
+      pattern: "*.filteringStats.tsv"
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+
+authors:
+  - "@GCJMackenzie"

modules/gatk4/getpileupsummaries/meta.yml

@@ -15,6 +15,7 @@ tools:
       homepage: https://gatk.broadinstitute.org/hc/en-us
       documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
       doi: 10.1158/1538-7445.AM2017-3590
+      licence: ['Apache-2.0']
 
 input:
   - meta:

modules/gatk4/haplotypecaller/meta.yml

@@ -13,6 +13,7 @@ tools:
     homepage: https://gatk.broadinstitute.org/hc/en-us
     documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
     doi: 10.1158/1538-7445.AM2017-3590
+    licence: ['Apache-2.0']
 
 input:
   - meta:

modules/gatk4/intervallisttools/meta.yml

@@ -14,6 +14,7 @@ tools:
       homepage: https://gatk.broadinstitute.org/hc/en-us
       documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
       doi: 10.1158/1538-7445.AM2017-3590
+      licence: ['Apache-2.0']
 
 input:
   - meta:

modules/gatk4/learnreadorientationmodel/meta.yml

@@ -15,6 +15,7 @@ tools:
       homepage: https://gatk.broadinstitute.org/hc/en-us
       documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
       doi: 10.1158/1538-7445.AM2017-3590
+      licence: ['Apache-2.0']
 
 input:
   - meta:

modules/gatk4/markduplicates/meta.yml

@@ -13,7 +13,7 @@ tools:
       documentation: https://gatk.broadinstitute.org/hc/en-us/articles/360037052812-MarkDuplicates-Picard-
       tool_dev_url: https://github.com/broadinstitute/gatk
       doi: 10.1158/1538-7445.AM2017-3590
-      licence: ['BSD-3-clause']
+      licence: ['MIT']
 
 input:
   - meta:

modules/gatk4/mergebamalignment/meta.yml

@@ -12,6 +12,7 @@ tools:
       homepage: https://gatk.broadinstitute.org/hc/en-us
       documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
       doi: 10.1158/1538-7445.AM2017-3590
+      licence: ['Apache-2.0']
 input:
   - meta:
       type: map

modules/gatk4/mergevcfs/meta.yml

@@ -12,6 +12,7 @@ tools:
       homepage: https://gatk.broadinstitute.org/hc/en-us
       documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
       doi: 10.1158/1538-7445.AM2017-3590
+      licence: ['Apache-2.0']
 input:
   - meta:
       type: map

modules/gatk4/mutect2/meta.yml

@@ -14,6 +14,7 @@ tools:
       homepage: https://gatk.broadinstitute.org/hc/en-us
       documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
       doi: 10.1158/1538-7445.AM2017-3590
+      licence: ['Apache-2.0']
 
 input:
   - meta:

modules/gatk4/revertsam/meta.yml

@@ -12,6 +12,7 @@ tools:
       homepage: https://gatk.broadinstitute.org/hc/en-us
       documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
       doi: 10.1158/1538-7445.AM2017-3590
+      licence: ['Apache-2.0']
 input:
   - meta:
       type: map

modules/gatk4/samtofastq/meta.yml

@@ -12,6 +12,7 @@ tools:
       homepage: https://gatk.broadinstitute.org/hc/en-us
       documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
       doi: 10.1158/1538-7445.AM2017-3590
+      licence: ['Apache-2.0']
 input:
   - meta:
       type: map

modules/gatk4/splitncigarreads/meta.yml

@@ -12,6 +12,7 @@ tools:
       homepage: https://gatk.broadinstitute.org/hc/en-us
       documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
       doi: 10.1158/1538-7445.AM2017-3590
+      licence: ['Apache-2.0']
 input:
   - meta:
       type: map

modules/gatk4/variantfiltration/meta.yml

@@ -12,6 +12,7 @@ tools:
       homepage: https://gatk.broadinstitute.org/hc/en-us
       documentation: https://gatk.broadinstitute.org/hc/en-us/categories/360002369672s
       doi: 10.1158/1538-7445.AM2017-3590
+      licence: ['Apache-2.0']
 input:
   - meta:
       type: map

modules/genmap/index/meta.yml

@@ -9,7 +9,7 @@ tools:
       documentation: https://github.com/cpockrandt/genmap
       tool_dev_url: https://github.com/cpockrandt/genmap
       doi: "10.1093/bioinformatics/btaa222"
-      licence: ['BSD']
+      licence: ['BSD-3-Clause']
 
 input:
   - fasta:

modules/genmap/mappability/meta.yml

@@ -9,7 +9,7 @@ tools:
       documentation: https://github.com/cpockrandt/genmap
       tool_dev_url: https://github.com/cpockrandt/genmap
       doi: "10.1093/bioinformatics/btaa222"
-      licence: ['BSD']
+      licence: ['BSD-3-Clause']
 
 input:
   - fasta:

modules/genrich/functions.nf

@@ -0,0 +1,78 @@
+//
+//  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()
+}
+
+//
+// Extract name of module from process name using $task.process
+//
+def getProcessName(task_process) {
+    return task_process.tokenize(':')[-1]
+}
+
+//
+// Function to initialise default values and to generate a Groovy Map of available options for nf-core modules
+//
+def initOptions(Map args) {
+    def Map options = [:]
+    options.args            = args.args ?: ''
+    options.args2           = args.args2 ?: ''
+    options.args3           = args.args3 ?: ''
+    options.publish_by_meta = args.publish_by_meta ?: []
+    options.publish_dir     = args.publish_dir ?: ''
+    options.publish_files   = args.publish_files
+    options.suffix          = args.suffix ?: ''
+    return options
+}
+
+//
+// Tidy up and join elements of a list to return a path string
+//
+def getPathFromList(path_list) {
+    def paths = path_list.findAll { item -> !item?.trim().isEmpty() }      // Remove empty entries
+    paths     = paths.collect { it.trim().replaceAll("^[/]+|[/]+\$", "") } // Trim whitespace and trailing slashes
+    return paths.join('/')
+}
+
+//
+// Function to save/publish module results
+//
+def saveFiles(Map args) {
+    def ioptions  = initOptions(args.options)
+    def path_list = [ ioptions.publish_dir ?: args.publish_dir ]
+
+    // Do not publish versions.yml unless running from pytest workflow
+    if (args.filename.equals('versions.yml') && !System.getenv("NF_CORE_MODULES_TEST")) {
+        return null
+    }
+    if (ioptions.publish_by_meta) {
+        def key_list = ioptions.publish_by_meta instanceof List ? ioptions.publish_by_meta : args.publish_by_meta
+        for (key in key_list) {
+            if (args.meta && key instanceof String) {
+                def path = key
+                if (args.meta.containsKey(key)) {
+                    path = args.meta[key] instanceof Boolean ? "${key}_${args.meta[key]}".toString() : args.meta[key]
+                }
+                path = path instanceof String ? path : ''
+                path_list.add(path)
+            }
+        }
+    }
+    if (ioptions.publish_files instanceof Map) {
+        for (ext in ioptions.publish_files) {
+            if (args.filename.endsWith(ext.key)) {
+                def ext_list = path_list.collect()
+                ext_list.add(ext.value)
+                return "${getPathFromList(ext_list)}/$args.filename"
+            }
+        }
+    } else if (ioptions.publish_files == null) {
+        return "${getPathFromList(path_list)}/$args.filename"
+    }
+}

modules/genrich/main.nf

@@ -0,0 +1,69 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName; getProcessName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process GENRICH {
+    tag "$meta.id"
+    label 'process_high'
+    publishDir "${params.outdir}",
+        mode: params.publish_dir_mode,
+        saveAs: { filename -> saveFiles(filename:filename, options:params.options, publish_dir:getSoftwareName(task.process), meta:meta, publish_by_meta:['id']) }
+
+    conda (params.enable_conda ? "bioconda::genrich=0.6.1" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/genrich:0.6.1--h5bf99c6_1"
+    } else {
+        container "quay.io/biocontainers/genrich:0.6.1--h5bf99c6_1"
+    }
+
+    input:
+    tuple val(meta), path(treatment_bam)
+    path  control_bam
+    path  blacklist_bed
+
+    output:
+    tuple val(meta), path("*narrowPeak")                     , emit: peaks
+    tuple val(meta), path("*pvalues.bedGraph"), optional:true, emit: bedgraph_pvalues
+    tuple val(meta), path("*pileup.bedGraph") , optional:true, emit: bedgraph_pileup
+    tuple val(meta), path("*intervals.bed")   , optional:true, emit: bed_intervals
+    tuple val(meta), path("*duplicates.txt")  , optional:true, emit: duplicates
+    path "versions.yml"                                      , emit: versions
+
+    script:
+    def prefix     = options.suffix              ? "${meta.id}${options.suffix}"   : "${meta.id}"
+    def control    = params.control_bam          ? "-c $control_bam"               : ''
+    def pvalues    = params.pvalues              ? "-f ${prefix}.pvalues.bedGraph" : ""
+    def pileup     = params.pileup               ? "-k ${prefix}.pileup.bedGraph"  : ""
+    def bed        = params.bed                  ? "-b ${prefix}.intervals.bed"    : ""
+    def blacklist  = params.blacklist_bed        ? "-E $blacklist_bed"             : ""
+    def duplicates = ""
+    if (params.save_duplicates) {
+        if (options.args.contains('-r')) {
+            duplicates = "-R ${prefix}.duplicates.txt"
+        } else {
+            log.info '[Genrich] Duplicates can only be saved if they are filtered, defaulting to -r option (Remove PCR duplicates).'
+            duplicates = "-r -R ${prefix}.duplicates.txt"
+        }
+    }
+    """
+    Genrich \\
+        -t $treatment_bam \\
+        $options.args \\
+        $control \\
+        $blacklist \\
+        -o ${prefix}.narrowPeak \\
+        $pvalues \\
+        $pileup \\
+        $bed \\
+        $duplicates \\
+        $blacklist \\
+        $control
+
+    cat <<-END_VERSIONS > versions.yml
+    ${getProcessName(task.process)}:
+        ${getSoftwareName(task.process)}: \$(echo \$(Genrich --version 2>&1) | sed 's/^Genrich, version //; s/ .*\$//')
+    END_VERSIONS
+    """
+}

modules/genrich/meta.yml

@@ -0,0 +1,71 @@
+name: genrich
+description: Peak-calling for ChIP-seq and ATAC-seq enrichment experiments
+keywords:
+  - peak-calling
+  - ChIP-seq
+  - ATAC-seq
+tools:
+  - genrich:
+      description: |
+        Genrich is a peak-caller for genomic enrichment assays (e.g. ChIP-seq, ATAC-seq).
+        It analyzes alignment files generated following the assay and produces a file
+        detailing peaks of significant enrichment.
+      homepage: https://github.com/jsh58/Genrich
+      documentation: https://github.com/jsh58/Genrich#readme
+      tool_dev_url: https://github.com/jsh58/Genrich
+      doi: ""
+      licence: ['MIT']
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - treatment_bam:
+      type: file
+      description: Coordinate sorted BAM/SAM file from treatment sample
+      pattern: "*.{bam,sam}"
+  - control_bam:
+      type: file
+      description: Coordinate sorted BAM/SAM file from control sample
+      pattern: "*.{bam,sam}"
+  - blacklist_bed:
+      type: file
+      description: Bed file containing genomic intervals to exclude from the analysis
+      pattern: "*.{bed}"
+
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - peaks:
+      type: file
+      description: Output file is in ENCODE narrowPeak format
+      pattern: "*.{narrowPeak}"
+  - bedgraph_pvalues:
+      type: file
+      description: bedGraph file containing p/q values
+      pattern: "*.{pvalues.bedGraph}"
+  - bedgraph_pileup:
+      type: file
+      description: bedGraph file containing pileups and p-values
+      pattern: "*.{pileup.bedGraph}"
+  - bed_intervals:
+      type: file
+      description: Bed file containing annotated intervals
+      pattern: "*.{intervals.bed}"
+  - duplicates:
+      type: file
+      description: Text output file containing intervals corresponding to PCR duplicates
+      pattern: "*.{intervals.txt}"
+  - version:
+      type: file
+      description: File containing software version
+      pattern: "*.{version.txt}"
+
+authors:
+  - "@JoseEspinosa"
+

modules/glnexus/meta.yml

@@ -10,7 +10,7 @@ tools:
       documentation: https://github.com/dnanexus-rnd/GLnexus/wiki/Getting-Started
       tool_dev_url: None
       doi: https://doi.org/10.1101/343970
-      licence: ['Apache License 2.0']
+      licence: ['Apache-2.0']
 
 input:
   - meta:

modules/graphmap2/align/meta.yml

@@ -12,6 +12,7 @@ tools:
             A versatile pairwise aligner for genomic and spliced nucleotide sequences.
         homepage: https://github.com/lbcb-sci/graphmap2
         documentation: https://github.com/lbcb-sci/graphmap2#graphmap2---a-highly-sensitive-and-accurate-mapper-for-long-error-prone-reads
+        licence: ['MIT']
 input:
     - meta:
         type: map

modules/graphmap2/index/meta.yml

@@ -10,6 +10,7 @@ tools:
             A versatile pairwise aligner for genomic and spliced nucleotide sequences.
         homepage: https://github.com/lbcb-sci/graphmap2
         documentation: https://github.com/lbcb-sci/graphmap2#graphmap2---a-highly-sensitive-and-accurate-mapper-for-long-error-prone-reads
+        licence: ['MIT']
 input:
     - fasta:
         type: file

modules/gstama/collapse/functions.nf

@@ -0,0 +1,78 @@
+//
+//  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()
+}
+
+//
+// Extract name of module from process name using $task.process
+//
+def getProcessName(task_process) {
+    return task_process.tokenize(':')[-1]
+}
+
+//
+// Function to initialise default values and to generate a Groovy Map of available options for nf-core modules
+//
+def initOptions(Map args) {
+    def Map options = [:]
+    options.args            = args.args ?: ''
+    options.args2           = args.args2 ?: ''
+    options.args3           = args.args3 ?: ''
+    options.publish_by_meta = args.publish_by_meta ?: []
+    options.publish_dir     = args.publish_dir ?: ''
+    options.publish_files   = args.publish_files
+    options.suffix          = args.suffix ?: ''
+    return options
+}
+
+//
+// Tidy up and join elements of a list to return a path string
+//
+def getPathFromList(path_list) {
+    def paths = path_list.findAll { item -> !item?.trim().isEmpty() }      // Remove empty entries
+    paths     = paths.collect { it.trim().replaceAll("^[/]+|[/]+\$", "") } // Trim whitespace and trailing slashes
+    return paths.join('/')
+}
+
+//
+// Function to save/publish module results
+//
+def saveFiles(Map args) {
+    def ioptions  = initOptions(args.options)
+    def path_list = [ ioptions.publish_dir ?: args.publish_dir ]
+
+    // Do not publish versions.yml unless running from pytest workflow
+    if (args.filename.equals('versions.yml') && !System.getenv("NF_CORE_MODULES_TEST")) {
+        return null
+    }
+    if (ioptions.publish_by_meta) {
+        def key_list = ioptions.publish_by_meta instanceof List ? ioptions.publish_by_meta : args.publish_by_meta
+        for (key in key_list) {
+            if (args.meta && key instanceof String) {
+                def path = key
+                if (args.meta.containsKey(key)) {
+                    path = args.meta[key] instanceof Boolean ? "${key}_${args.meta[key]}".toString() : args.meta[key]
+                }
+                path = path instanceof String ? path : ''
+                path_list.add(path)
+            }
+        }
+    }
+    if (ioptions.publish_files instanceof Map) {
+        for (ext in ioptions.publish_files) {
+            if (args.filename.endsWith(ext.key)) {
+                def ext_list = path_list.collect()
+                ext_list.add(ext.value)
+                return "${getPathFromList(ext_list)}/$args.filename"
+            }
+        }
+    } else if (ioptions.publish_files == null) {
+        return "${getPathFromList(path_list)}/$args.filename"
+    }
+}

modules/gstama/collapse/main.nf

@@ -0,0 +1,52 @@
+// Import generic module functions
+include { initOptions; saveFiles; getSoftwareName; getProcessName } from './functions'
+
+params.options = [:]
+options        = initOptions(params.options)
+
+process GSTAMA_COLLAPSE {
+    tag "$meta.id"
+    label 'process_medium'
+    publishDir "${params.outdir}",
+        mode: params.publish_dir_mode,
+        saveAs: { filename -> saveFiles(filename:filename, options:params.options, publish_dir:getSoftwareName(task.process), meta:meta, publish_by_meta:['id']) }
+
+    conda (params.enable_conda ? "bioconda::gs-tama=1.0.2" : null)
+    if (workflow.containerEngine == 'singularity' && !params.singularity_pull_docker_container) {
+        container "https://depot.galaxyproject.org/singularity/gs-tama:1.0.2--hdfd78af_0"
+    } else {
+        container "quay.io/biocontainers/gs-tama:1.0.2--hdfd78af_0"
+    }
+
+    input:
+    tuple val(meta), path(bam)
+    path fasta
+
+    output:
+    tuple val(meta), path("*.bed")                    , emit: bed
+    tuple val(meta), path("*_trans_read.bed")         , emit: bed_trans_reads
+    tuple val(meta), path("*_local_density_error.txt"), emit: local_density_error
+    tuple val(meta), path("*_polya.txt")              , emit: polya
+    tuple val(meta), path("*_read.txt")               , emit: read
+    tuple val(meta), path("*_strand_check.txt")       , emit: strand_check
+    tuple val(meta), path("*_trans_report.txt")       , emit: trans_report
+    path "versions.yml"                               , emit: versions
+
+    tuple val(meta), path("*_varcov.txt")             , emit: varcov  , optional: true
+    tuple val(meta), path("*_variants.txt")           , emit: variants, optional: true
+
+    script:
+    def prefix = options.suffix ? "${meta.id}${options.suffix}" : "${meta.id}"
+    """
+    tama_collapse.py \\
+        -s $bam \\
+        -f $fasta \\
+        -p ${prefix} \\
+        $options.args
+
+    cat <<-END_VERSIONS > versions.yml
+    ${getProcessName(task.process)}:
+        ${getSoftwareName(task.process)}: \$( tama_collapse.py -version | grep 'tc_version_date_'|sed 's/tc_version_date_//g' )
+    END_VERSIONS
+    """
+}