Merge pull request #1620 from mahesh-panchal/kat_hist

Add module kat hist

modules/kat/hist/main.nf

@@ -0,0 +1,42 @@
+process KAT_HIST {
+    tag "$meta.id"
+    label 'process_medium'
+
+    conda (params.enable_conda ? "bioconda::kat=2.4.2" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/kat:2.4.2--py38hfc5f9d8_2':
+        'quay.io/biocontainers/kat:2.4.2--py38hfc5f9d8_2' }"
+
+    input:
+    tuple val(meta), path(reads)
+
+    output:
+    tuple val(meta), path("*.hist")                   , emit: hist
+    tuple val(meta), path("*.hist.dist_analysis.json"), emit: json
+    tuple val(meta), path("*.png")                    , emit: png           , optional: true
+    tuple val(meta), path("*.ps")                     , emit: ps            , optional: true
+    tuple val(meta), path("*.pdf")                    , emit: pdf           , optional: true
+    tuple val(meta), path("*-hash.jf*")               , emit: jellyfish_hash, 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}"
+    """
+    kat hist \\
+        --threads $task.cpus \\
+        --output_prefix ${prefix}.hist \\
+        $args \\
+        $reads
+
+    ls -l
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        kat: \$( kat hist --version | sed 's/kat //' )
+    END_VERSIONS
+    """
+}

modules/kat/hist/meta.yml

@@ -0,0 +1,64 @@
+name: "kat_hist"
+description: Creates a histogram of the number of distinct k-mers having a given frequency.
+keywords:
+  - k-mer
+  - histogram
+  - count
+tools:
+  - "kat":
+      description: "KAT is a suite of tools that analyse jellyfish hashes or sequence files (fasta or fastq) using kmer counts"
+      homepage: https://www.earlham.ac.uk/kat-tools
+      documentation: https://kat.readthedocs.io/en/latest/index.html
+      tool_dev_url: https://github.com/TGAC/KAT
+      doi: http://bioinformatics.oxfordjournals.org/content/early/2016/10/20/bioinformatics.btw663.abstract
+      licence: "['GPL v3']"
+
+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"
+  - hist:
+      type: file
+      description: KAT histogram of k-mer counts
+      pattern: "*.hist"
+  - json:
+      type: file
+      description: KAT histogram summary of distance analysis
+      pattern: "*.hist.dist_analysis.json"
+  - png:
+      type: file
+      description: KAT plot of k-mer histogram in PNG format
+      pattern: "*.png"
+  - ps:
+      type: file
+      description: KAT plot of k-mer histogram in PS format
+      pattern: "*.ps"
+  - pdf:
+      type: file
+      description: KAT plot of k-mer histogram in PDF format
+      pattern: "*.pdf"
+  - jellyfish_hash:
+      type: file
+      description: Jellyfish hash file
+      pattern: "*-hist.jf*"
+
+authors:
+  - "@mahesh-panchal"

tests/config/pytest_modules.yml

@@ -1093,6 +1093,10 @@ kallistobustools/ref:
   - modules/kallistobustools/ref/**
   - tests/modules/kallistobustools/ref/**
 
+kat/hist:
+  - modules/kat/hist/**
+  - tests/modules/kat/hist/**
+
 khmer/normalizebymedian:
   - modules/khmer/normalizebymedian/**
   - tests/modules/khmer/normalizebymedian/**

tests/modules/kat/hist/main.nf

@@ -0,0 +1,28 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { KAT_HIST } from '../../../../modules/kat/hist/main.nf'
+
+workflow test_kat_hist_single_end {
+
+    input = [
+        [ id:'test', single_end:true ], // meta map
+        file(params.test_data['homo_sapiens']['illumina']['test2_1_fastq_gz'], checkIfExists: true)
+    ]
+
+    KAT_HIST ( input )
+}
+
+workflow test_kat_hist_paired_end {
+
+    input = [
+        [ id:'test', single_end:false ], // meta map
+        [
+            file(params.test_data['homo_sapiens']['illumina']['test2_1_fastq_gz'], checkIfExists: true),
+            file(params.test_data['homo_sapiens']['illumina']['test2_2_fastq_gz'], checkIfExists: true),
+        ]
+    ]
+
+    KAT_HIST ( input )
+}

tests/modules/kat/hist/nextflow.config

@@ -0,0 +1,9 @@
+process {
+
+    publishDir = { "${params.outdir}/${task.process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()}" }
+
+    withName: 'test_kat_hist_single_end:KAT_HIST' {
+        ext.args = '-d'
+    }
+
+}

tests/modules/kat/hist/test.yml

@@ -0,0 +1,42 @@
+- name: kat hist test_kat_hist_single_end
+  command: nextflow run tests/modules/kat/hist -entry test_kat_hist_single_end -c tests/config/nextflow.config
+  tags:
+    - kat/hist
+    - kat
+  files:
+    - path: output/kat/test.hist
+      md5sum: c6eba52b3a2653a684577a8ae20b74c1
+    - path: output/kat/test.hist-hash.jf27
+    - path: output/kat/test.hist.dist_analysis.json
+      # md5sum: 52a5a2d91c71b940f36f1f0a7fd5ef10 # This is variable for an unknown reason
+      contains:
+        - "nb_peaks"
+        - "global_minima"
+        - "global_maxima"
+        - "mean_freq"
+        - "est_genome_size"
+        - "est_het_rate"
+    - path: output/kat/test.hist.png
+      md5sum: 49861ef1a265e0edde3550b39c64a274
+    - path: output/kat/versions.yml
+
+- name: kat hist test_kat_hist_paired_end
+  command: nextflow run tests/modules/kat/hist -entry test_kat_hist_paired_end -c tests/config/nextflow.config
+  tags:
+    - kat/hist
+    - kat
+  files:
+    - path: output/kat/test.hist
+      md5sum: 91429091e74b1718051591d83a1ccb5d
+    - path: output/kat/test.hist.dist_analysis.json
+      # md5sum: 8b0dabeaff4ba706b33aa8964d687e13  # This is variable for an unknown reason
+      contains:
+        - "nb_peaks"
+        - "global_minima"
+        - "global_maxima"
+        - "mean_freq"
+        - "est_genome_size"
+        - "est_het_rate"
+    - path: output/kat/test.hist.png
+      md5sum: e20774d0d2b979cb6ead7b7fb5ad36d9
+    - path: output/kat/versions.yml