Dereplicate

modules/nf-core/vsearch/dereplicate/environment.yml

@@ -0,0 +1,7 @@
+name: "vsearch_dereplicate"
+channels:
+  - conda-forge
+  - bioconda
+  - defaults
+dependencies:
+  - "bioconda::vsearch=2.28.1"

modules/nf-core/vsearch/dereplicate/main.nf

@@ -0,0 +1,52 @@
+process VSEARCH_DEREPLICATE {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/vsearch:2.28.1--h6a68c12_1':
+        'biocontainers/vsearch:2.28.1--h6a68c12_1' }"
+
+    input:
+    tuple val(meta), path(fasta)
+
+    output:
+    tuple val(meta), path('*.derep.fasta')   , emit: fasta
+    tuple val(meta), path('*.derep.uc')      , emit: clustering
+    path "*.derep.log"                       , emit: log
+    path "versions.yml"                      , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    vsearch \\
+        --derep_fulllength ${fasta} \\
+        $args \\
+        --relabel "${prefix}." \\
+        --uc ${prefix}.derep.uc \\
+        --output ${prefix}.derep.fasta 2>&1 | tee ${prefix}.derep.log
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        vsearch: \$(vsearch --version 2>&1 | head -n 1 | sed 's/vsearch //g' | sed 's/,.*//g' | sed 's/^v//' | sed 's/_.*//')
+    END_VERSIONS
+    """
+
+    stub:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    touch ${prefix}.derep.fasta
+    touch ${prefix}.derep.uc
+    touch myfile.derep.log
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        vsearch: \$(vsearch --version 2>&1 | head -n 1 | sed 's/vsearch //g' | sed 's/,.*//g' | sed 's/^v//' | sed 's/_.*//')
+    END_VERSIONS
+    """
+}

modules/nf-core/vsearch/dereplicate/meta.yml

@@ -0,0 +1,50 @@
+name: "vsearch_dereplicate"
+description: Merge strictly identical sequences contained in filename. Identical sequences are defined as having the same length and the same string of nucleotides (case insensitive, T and U are considered the same).
+keywords:
+  - vsearch/dereplicate
+  - vsearch
+  - dereplicate
+  - amplicon sequences
+  - metagenomics
+  - genomics
+  - population genetics
+tools:
+  - "vsearch":
+      description: "A versatile open source tool for metagenomics (USEARCH alternative)"
+      homepage: https://github.com/torognes/vsearch
+      documentation: https://github.com/torognes/vsearch/releases/download/v2.21.1/vsearch_manual.pdf
+      tool_dev_url: https://github.com/torognes/vsearch
+      doi: 10.7717/peerj.2584
+      licence: ["GPL v3-or-later OR BSD-2-clause"]
+input:
+  - meta:
+      type: map
+      description: Groovy Map containing sample information e.g. [ id:'test' ]
+  - fasta:
+      type: file
+      description: Sequences to be sorted in FASTA format
+      pattern: "*.{fasta,fa,fasta.gz,fa.gz,.fna,.fna.gz}"
+output:
+  - meta:
+      type: map
+      description: Groovy Map containing sample information e.g. [ id:'test' ]`
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+  - fasta:
+      type: file
+      description: dereplicated fasta
+      pattern: "*.fasta"
+  - clustering:
+      type: file
+      description: dereplicated derep.uc file
+      pattern: "*.uc"
+  - log:
+      type: file
+      description: a log file of the run
+      pattern: "*.derep.log"
+authors:
+  - "@chriswyatt1"
+maintainers:
+  - "@chriswyatt1"

modules/nf-core/vsearch/dereplicate/tests/main.nf.test

@@ -0,0 +1,58 @@
+nextflow_process {
+
+    name "Test Process VSEARCH_DEREPLICATE"
+    script "../main.nf"
+    process "VSEARCH_DEREPLICATE"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "vsearch"
+    tag "vsearch/dereplicate"
+
+    test("phyloplace - fasta") {
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ "id": "ERR2537816", "single_end": false ],
+                    file("https://raw.githubusercontent.com/nf-core/test-datasets/phyloplace/testdata/cyn_syn.fna", checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+    test("phyloplace -fasta - stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ "id": "ERR2537816", "single_end": false ],
+                    file("https://raw.githubusercontent.com/nf-core/test-datasets/phyloplace/testdata/cyn_syn.fna", checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+}

modules/nf-core/vsearch/dereplicate/tests/main.nf.test.snap

@@ -0,0 +1,120 @@
+{
+    "pollon -fasta - stub": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "ERR2537816",
+                            "single_end": false
+                        },
+                        "ERR2537816.derep.fasta:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "1": [
+                    [
+                        {
+                            "id": "ERR2537816",
+                            "single_end": false
+                        },
+                        "ERR2537816.derep.uc:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "2": [
+                    "myfile.derep.log:md5,d41d8cd98f00b204e9800998ecf8427e"
+                ],
+                "3": [
+                    "versions.yml:md5,b00eb984d2fdee20b07a68bfc9fa08cb"
+                ],
+                "clustering": [
+                    [
+                        {
+                            "id": "ERR2537816",
+                            "single_end": false
+                        },
+                        "ERR2537816.derep.uc:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "fasta": [
+                    [
+                        {
+                            "id": "ERR2537816",
+                            "single_end": false
+                        },
+                        "ERR2537816.derep.fasta:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "log": [
+                    "myfile.derep.log:md5,d41d8cd98f00b204e9800998ecf8427e"
+                ],
+                "versions": [
+                    "versions.yml:md5,b00eb984d2fdee20b07a68bfc9fa08cb"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.0",
+            "nextflow": "24.04.4"
+        },
+        "timestamp": "2024-08-24T06:58:16.351579571"
+    },
+    "pollon - fasta": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "ERR2537816",
+                            "single_end": false
+                        },
+                        "ERR2537816.derep.fasta:md5,04313957d264c2fd5cc067920af51b08"
+                    ]
+                ],
+                "1": [
+                    [
+                        {
+                            "id": "ERR2537816",
+                            "single_end": false
+                        },
+                        "ERR2537816.derep.uc:md5,99b94a34c5e2ba5fd37bff7c4b46aca8"
+                    ]
+                ],
+                "2": [
+                    "ERR2537816.derep.log:md5,fe97ec773766190003f0e53cda8a90f9"
+                ],
+                "3": [
+                    "versions.yml:md5,b00eb984d2fdee20b07a68bfc9fa08cb"
+                ],
+                "clustering": [
+                    [
+                        {
+                            "id": "ERR2537816",
+                            "single_end": false
+                        },
+                        "ERR2537816.derep.uc:md5,99b94a34c5e2ba5fd37bff7c4b46aca8"
+                    ]
+                ],
+                "fasta": [
+                    [
+                        {
+                            "id": "ERR2537816",
+                            "single_end": false
+                        },
+                        "ERR2537816.derep.fasta:md5,04313957d264c2fd5cc067920af51b08"
+                    ]
+                ],
+                "log": [
+                    "ERR2537816.derep.log:md5,fe97ec773766190003f0e53cda8a90f9"
+                ],
+                "versions": [
+                    "versions.yml:md5,b00eb984d2fdee20b07a68bfc9fa08cb"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.0",
+            "nextflow": "24.04.4"
+        },
+        "timestamp": "2024-08-24T06:58:07.324653651"
+    }
+}