New module: khmer/trimlowabund

modules/nf-core/khmer/trimlowabund/environment.yml

@@ -0,0 +1,7 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/environment-schema.json
+channels:
+  - conda-forge
+  - bioconda
+dependencies:
+  - "bioconda::khmer=3.0.0a3"

modules/nf-core/khmer/trimlowabund/main.nf

@@ -0,0 +1,49 @@
+process KHMER_TRIMLOWABUND {
+    tag "${meta.id}"
+    label 'process_low'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/khmer:3.0.0a3--py37haa7609a_2' :
+        'biocontainers/khmer:3.0.0a3--py37haa7609a_2' }"
+
+    input:
+    tuple val(meta), path(fasta)
+
+    output:
+    tuple val(meta), path("*.abundtrim"), emit: sequence
+    path "versions.yml"                 , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    if (!task.memory) {
+        log.info '[KHMER_TRIMLOWABUND] Available memory not known - defaulting to 16GB. Specify process memory requirements to change this.'
+        avail_mem = 16
+    } else {
+        avail_mem = task.memory.toGiga()
+    }
+    """
+    trim-low-abund.py \\
+        -M ${avail_mem}e9 \\
+        $args \\
+        $fasta
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        khmer: \$( trim-low-abund.py --version 2>&1 | grep ^khmer | sed 's/^khmer //' )
+    END_VERSIONS
+    """
+
+    stub:
+    """
+    touch ${fasta}.abundtrim
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        khmer: \$( trim-low-abund.py --version 2>&1 | grep ^khmer | sed 's/^khmer //' )
+    END_VERSIONS
+    """
+}

modules/nf-core/khmer/trimlowabund/meta.yml

@@ -0,0 +1,51 @@
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/meta-schema.json
+name: "khmer_trimlowabund"
+description: Removes low abundance k-mers from reads
+keywords:
+  - quality control
+  - genomics
+  - filtering
+  - reads
+  - khmer
+  - k-mer
+tools:
+  - "khmer":
+      description: khmer k-mer counting library
+      homepage: https://github.com/dib-lab/khmer
+      documentation: https://khmer.readthedocs.io/en/latest/
+      tool_dev_url: https://github.com/dib-lab/khmer
+      doi: "10.12688/f1000research.6924.1"
+      licence: ["BSD License"]
+      identifier: biotools:khmer
+input:
+  - - meta:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'sample1', single_end:false ]`
+    - fasta:
+        type: file
+        description: fasta file
+        pattern: "*.{fa,fasta,fas}"
+
+output:
+  - sequence:
+    - meta:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'sample1', single_end:false ]`
+    - '*.abundtrim':
+        type: file
+        description: fasta file with low abundance k-mers removed.
+        pattern: "*.{fa,fasta,fas}"
+  - versions:
+    - "versions.yml":
+        type: file
+        description: File containing software versions
+        pattern: "versions.yml"
+
+authors:
+  - "@zachary-foster"
+maintainers:
+  - "@zachary-foster"

modules/nf-core/khmer/trimlowabund/tests/main.nf.test

@@ -0,0 +1,58 @@
+nextflow_process {
+
+    name "Test Process KHMER_TRIMLOWABUND"
+    script "../main.nf"
+    process "KHMER_TRIMLOWABUND"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "khmer"
+    tag "khmer/trimlowabund"
+
+    test("sarscov2 - fastq.gz") {
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id:'test', single_end:false ], // meta map
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test_2.fastq.gz', checkIfExists: true),
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+    test("sarscov2 - fastq.gz - stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id:'test', single_end:false ], // meta map
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test_2.fastq.gz', checkIfExists: true),
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+}

modules/nf-core/khmer/trimlowabund/tests/main.nf.test.snap

@@ -0,0 +1,72 @@
+{
+    "sarscov2 - fastq.gz": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test",
+                            "single_end": false
+                        },
+                        "test_2.fastq.gz.abundtrim:md5,36716902820bac37e2a10e99e2c94c88"
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,867953e387d894675fbac67fbb05cb2a"
+                ],
+                "sequence": [
+                    [
+                        {
+                            "id": "test",
+                            "single_end": false
+                        },
+                        "test_2.fastq.gz.abundtrim:md5,36716902820bac37e2a10e99e2c94c88"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,867953e387d894675fbac67fbb05cb2a"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.5"
+        },
+        "timestamp": "2025-03-14T15:23:28.692050011"
+    },
+    "sarscov2 - fastq.gz - stub": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test",
+                            "single_end": false
+                        },
+                        "test_2.fastq.gz.abundtrim:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,867953e387d894675fbac67fbb05cb2a"
+                ],
+                "sequence": [
+                    [
+                        {
+                            "id": "test",
+                            "single_end": false
+                        },
+                        "test_2.fastq.gz.abundtrim:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,867953e387d894675fbac67fbb05cb2a"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.5"
+        },
+        "timestamp": "2025-03-14T15:23:34.05055035"
+    }
+}