FastK/fastK (#1739)

* Add FastK/FastK

* Add ktab and prof outputs

* Add the hidden files necessary for the optional files

* Update image

* Apply suggestions from code review

Co-authored-by: Daniel Straub <42973691+d4straub@users.noreply.github.com>

Co-authored-by: Daniel Straub <42973691+d4straub@users.noreply.github.com>

modules/fastk/fastk/main.nf

@@ -0,0 +1,38 @@
+process FASTK_FASTK {
+    tag "$meta.id"
+    label 'process_medium'
+
+    if (params.enable_conda) {
+        error "Conda environments cannot be used when using the FastK tool. Please use docker or singularity containers."
+    }
+    container 'ghcr.io/nbisweden/fastk_genescopefk_merquryfk:1.0'
+
+    input:
+    tuple val(meta), path(reads)
+
+    output:
+    tuple val(meta), path("*.hist")               , emit: hist
+    tuple val(meta), path("*.ktab*", hidden: true), emit: ktab, optional: true
+    tuple val(meta), path("*.prof*", hidden: true), emit: prof, optional: true
+    path "versions.yml"                           , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    def FASTK_VERSION = 'f18a4e6d2207539f7b84461daebc54530a9559b0'
+    """
+    FastK \\
+        $args \\
+        -T$task.cpus \\
+        -N${prefix}_fk \\
+        $reads
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        fastk: $FASTK_VERSION
+    END_VERSIONS
+    """
+}

modules/fastk/fastk/meta.yml

@@ -0,0 +1,52 @@
+name: "fastk_fastk"
+description: A fast K-mer counter for high-fidelity shotgun datasets
+keywords:
+  - k-mer
+  - count
+  - histogram
+tools:
+  - "fastk":
+      description: "A fast K-mer counter for high-fidelity shotgun datasets"
+      homepage: "https://github.com/thegenemyers/FASTK"
+      documentation: ""
+      tool_dev_url: "https://github.com/thegenemyers/FASTK"
+      doi: ""
+      licence: "https://github.com/thegenemyers/FASTK/blob/master/LICENSE"
+
+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: Histogram of k-mers
+      pattern: "*.hist"
+  - ktab:
+      type: file
+      description: A sorted table of all canonical k‑mers along with their counts.
+      pattern: "*.ktab"
+  - prof:
+      type: file
+      description: A k‑mer count profile of each sequence in the input data set.
+      pattern: "*.prof"
+
+authors:
+  - "@mahesh-panchal"

tests/config/pytest_modules.yml

@@ -675,6 +675,10 @@ fastani:
   - modules/fastani/**
   - tests/modules/fastani/**
 
+fastk/fastk:
+  - modules/fastk/fastk/**
+  - tests/modules/fastk/fastk/**
+
 fastp:
   - modules/fastp/**
   - tests/modules/fastp/**

tests/modules/fastk/fastk/main.nf

@@ -0,0 +1,28 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { FASTK_FASTK } from '../../../../modules/fastk/fastk/main.nf'
+
+workflow test_fastk_fastk_single_end {
+
+    input = [
+        [ id:'test' , single_end: true ], // meta map
+        file(params.test_data['sarscov2']['illumina']['test_1_fastq_gz'], checkIfExists: true)
+    ]
+
+    FASTK_FASTK ( input )
+}
+
+workflow test_fastk_fastk_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)
+        ]
+    ]
+
+    FASTK_FASTK ( input )
+}

tests/modules/fastk/fastk/nextflow.config

@@ -0,0 +1,6 @@
+process {
+
+    publishDir = { "${params.outdir}/${task.process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()}" }
+    ext.args = '-t -p'
+
+}

tests/modules/fastk/fastk/test.yml

@@ -0,0 +1,39 @@
+- name: fastk fastk test_fastk_fastk_single_end
+  command: nextflow run ./tests/modules/fastk/fastk -entry test_fastk_fastk_single_end -c ./tests/config/nextflow.config  -c ./tests/modules/fastk/fastk/nextflow.config
+  tags:
+    - fastk
+    - fastk/fastk
+  files:
+    - path: output/fastk/.test_fk.ktab.1
+      md5sum: ceeacd0cb3aa69bf9b2a402830b40e26
+    - path: output/fastk/.test_fk.ktab.2
+      md5sum: f2629fd15b285aed3dc2d5fe546edf3f
+    - path: output/fastk/.test_fk.prof.1
+      md5sum: ebd48923a724cf79934f0b2ed42ba73d
+    - path: output/fastk/test_fk.hist
+      md5sum: c80e12f7321e62dba4b437d7bff36ec0
+    - path: output/fastk/test_fk.ktab
+      md5sum: a605a58931a4b5029469e1c2575c8cee
+    - path: output/fastk/test_fk.prof
+      md5sum: 43d426c95d277b8148406624d513bd40
+
+- name: fastk fastk test_fastk_fastk_paired_end
+  command: nextflow run ./tests/modules/fastk/fastk -entry test_fastk_fastk_paired_end -c ./tests/config/nextflow.config  -c ./tests/modules/fastk/fastk/nextflow.config
+  tags:
+    - fastk
+    - fastk/fastk
+  files:
+    - path: output/fastk/.test_fk.ktab.1
+      md5sum: 7f28fb44940fda799797e3069f5d7263
+    - path: output/fastk/.test_fk.ktab.2
+      md5sum: c14a85c128926ace78372f09029977b1
+    - path: output/fastk/.test_fk.prof.1
+      md5sum: 46a5fd9e297262b058f8c1fd062fcf56
+    - path: output/fastk/.test_fk.prof.2
+      md5sum: 80326a7406f41ccf2e51e341fc804132
+    - path: output/fastk/test_fk.hist
+      md5sum: 4f75b550d87ed4f26a2b10a05ac7e98c
+    - path: output/fastk/test_fk.ktab
+      md5sum: fddd5be0c36ad1d2131b8d8774f7657a
+    - path: output/fastk/test_fk.prof
+      md5sum: d3c7d8decd4ea6e298291b8be0e2de85