new module: fastaguard

modules/nf-core/fastaguard/environment.yml

@@ -0,0 +1,5 @@
+channels:
+  - conda-forge
+  - bioconda
+dependencies:
+  - bioconda::fastaguard=0.6.0

modules/nf-core/fastaguard/main.nf

@@ -0,0 +1,34 @@
+process FASTAGUARD {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine in ['singularity', 'apptainer'] && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/fastaguard:0.6.0--hfa8f182_0':
+        'quay.io/biocontainers/fastaguard:0.6.0--hfa8f182_0' }"
+
+    input:
+    tuple val(meta), path(fasta)
+
+    output:
+    tuple val(meta), path("*.fastaguard.html"), emit: html
+    tuple val(meta), path("*.fastaguard.json"), emit: json
+    tuple val(meta), path("*.fastaguard.tsv"), emit: tsv
+    tuple val(meta), path("*.fastaguard_mqc.json"), emit: mqc
+    tuple val("${task.process}"), val('fastaguard'), eval('fastaguard --version | cut -d " " -f 2'), emit: versions_fastaguard, topic: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    def args = task.ext.args ?: ''
+    """
+    fastaguard ${fasta} \
+      ${args} \
+      --out ${prefix}.fastaguard.html \
+      --json ${prefix}.fastaguard.json \
+      --tsv ${prefix}.fastaguard.tsv \
+      --multiqc ${prefix}.fastaguard_mqc.json
+    """
+}

modules/nf-core/fastaguard/meta.yml

@@ -0,0 +1,97 @@
+name: "fastaguard"
+description: FASTA preflight QC for assembly pipelines
+keywords:
+  - fasta
+  - assembly
+  - qc
+tools:
+  - "fastaguard":
+      description: FASTA preflight QC for assembly pipelines
+      homepage: "https://github.com/ehsanestaji/FastaGuard"
+      documentation: "https://github.com/ehsanestaji/FastaGuard"
+      tool_dev_url: "https://github.com/ehsanestaji/FastaGuard"
+      licence:
+        - "MIT"
+      identifier: ""
+input:
+  - - meta:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. [ id:'sample' ]
+    - fasta:
+        type: file
+        description: Assembly FASTA file
+        pattern: "*.{fa,fasta,fa.gz,fasta.gz}"
+        ontologies: []
+output:
+  html:
+    - - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. [ id:'sample' ]
+      - "*.fastaguard.html":
+          type: file
+          description: FastaGuard HTML report
+          pattern: "*.fastaguard.html"
+          ontologies: []
+  json:
+    - - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. [ id:'sample' ]
+      - "*.fastaguard.json":
+          type: file
+          description: FastaGuard JSON report
+          pattern: "*.fastaguard.json"
+          ontologies:
+            - edam: http://edamontology.org/format_3464
+  tsv:
+    - - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. [ id:'sample' ]
+      - "*.fastaguard.tsv":
+          type: file
+          description: FastaGuard TSV summary
+          pattern: "*.fastaguard.tsv"
+          ontologies:
+            - edam: http://edamontology.org/format_3475
+  mqc:
+    - - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. [ id:'sample' ]
+      - "*.fastaguard_mqc.json":
+          type: file
+          description: MultiQC custom-content JSON
+          pattern: "*.fastaguard_mqc.json"
+          ontologies:
+            - edam: http://edamontology.org/format_3464
+  versions_fastaguard:
+    - - ${task.process}:
+          type: string
+          description: The process the versions were collected from
+      - fastaguard:
+          type: string
+          description: The tool name
+      - fastaguard --version | cut -d " " -f 2:
+          type: eval
+          description: The expression to obtain the version of the tool
+topics:
+  versions:
+    - - ${task.process}:
+          type: string
+          description: The process the versions were collected from
+      - fastaguard:
+          type: string
+          description: The tool name
+      - fastaguard --version | cut -d " " -f 2:
+          type: eval
+          description: The expression to obtain the version of the tool
+authors:
+  - "@ehsanestaji"

modules/nf-core/fastaguard/tests/main.nf.test

@@ -0,0 +1,123 @@
+nextflow_process {
+
+    name "Test FASTAGUARD"
+    script "../main.nf"
+    process "FASTAGUARD"
+    config "./nextflow.config"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "fastaguard"
+    tag "fastaguard_single"
+
+    test("pass FASTA emits all reports") {
+        when {
+            process {
+                """
+                def fasta = file(workDir.resolve("pass.fa").toString())
+                fasta.text = [
+                    ">clean",
+                    ("ACGT" * 60)
+                ].join("\\n") + "\\n"
+
+                input[0] = [
+                    [ id:'pass' ],
+                    fasta
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(sanitizeOutput(process.out, unstableKeys: ["html", "json", "tsv", "mqc"])).match() }
+            )
+        }
+    }
+
+    test("warn FASTA preserves reports") {
+        when {
+            process {
+                """
+                def fasta = file(workDir.resolve("warn.fa").toString())
+                fasta.text = [
+                    ">long",
+                    ("ACGT" * 60),
+                    ">tiny",
+                    "ACGT"
+                ].join("\\n") + "\\n"
+
+                input[0] = [
+                    [ id:'warn' ],
+                    fasta
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(sanitizeOutput(process.out, unstableKeys: ["html", "json", "tsv", "mqc"])).match() }
+            )
+        }
+    }
+
+    test("fail FASTA preserves reports for gate review") {
+        when {
+            process {
+                """
+                def fasta = file(workDir.resolve("fail.fa").toString())
+                fasta.text = [
+                    ">dup",
+                    "ACGTACGT",
+                    ">dup",
+                    "ACGTACGT",
+                    ">bad",
+                    "ACGTXYZ"
+                ].join("\\n") + "\\n"
+
+                input[0] = [
+                    [ id:'fail' ],
+                    fasta
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(sanitizeOutput(process.out, unstableKeys: ["html", "json", "tsv", "mqc"])).match() }
+            )
+        }
+    }
+
+    test("invalid FASTA is represented in the evidence path") {
+        when {
+            process {
+                """
+                def fasta = file(workDir.resolve("invalid.fa").toString())
+                fasta.text = [
+                    ">empty_record",
+                    ">next_record",
+                    "ACGT"
+                ].join("\\n") + "\\n"
+
+                input[0] = [
+                    [ id:'invalid' ],
+                    fasta
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(sanitizeOutput(process.out, unstableKeys: ["html", "json", "tsv", "mqc"])).match() }
+            )
+        }
+    }
+}