Add new module: disambiguate

modules/nf-core/disambiguate/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::ngs-disambiguate=2018.05.03"

modules/nf-core/disambiguate/main.nf

@@ -0,0 +1,45 @@
+process DISAMBIGUATE {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/ngs-disambiguate:2018.05.03--h06902ac_10':
+        'biocontainers/ngs-disambiguate:2018.05.03--h06902ac_10' }"
+
+    input:
+    tuple val(meta), path(bam_a), path(bam_b)
+
+    output:
+    tuple val(meta), path("*disambiguatedSpeciesA.bam"),                                                      emit: disambiguated_bam_a
+    tuple val(meta), path("*disambiguatedSpeciesB.bam"),                                                      emit: disambiguated_bam_b
+    tuple val(meta), path("*ambiguousSpeciesA.bam"),                                                          emit: ambiguous_bam_a
+    tuple val(meta), path("*ambiguousSpeciesB.bam"),                                                          emit: ambiguous_bam_b
+    tuple val(meta), path("*_summary.txt"),                                                                   emit: summary
+    tuple val("${task.process}"), val('ngs-disambiguate'), val('2018.05.03'), topic: versions, emit: versions_ngs_disambiguate
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args   = task.ext.args   ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    ngs_disambiguate \\
+        ${bam_a} \\
+        ${bam_b} \\
+        $args \\
+        -s ${prefix} \\
+        -o .
+    """
+
+    stub:
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    touch ${prefix}.disambiguatedSpeciesA.bam
+    touch ${prefix}.disambiguatedSpeciesB.bam
+    touch ${prefix}.ambiguousSpeciesA.bam
+    touch ${prefix}.ambiguousSpeciesB.bam
+    touch ${prefix}_summary.txt
+    """
+}

modules/nf-core/disambiguate/meta.yml

@@ -0,0 +1,134 @@
+name: "disambiguate"
+description: |
+  Disambiguates reads aligned to two different organisms (e.g. human and mouse)
+  from the same source of FASTQ files. Useful in explant RNA/DNA-Seq workflows
+  where reads from two species are present. For reads aligned to both organisms,
+  the algorithm compares alignment quality scores to determine the most likely
+  species of origin. Produces four BAM files (uniquely assigned to species A or B,
+  ambiguous for species A or B) and a summary file.
+keywords:
+  - disambiguation
+  - xenograft
+  - explant
+  - rna-seq
+  - alignment
+  - bam
+tools:
+  - "ngs-disambiguate":
+      description: "Disambiguation algorithm for reads aligned to two different organisms\
+        \ using Tophat, Hisat2, BWA or STAR alignments."
+      homepage: "https://github.com/AstraZeneca-NGS/disambiguate"
+      documentation: "https://github.com/AstraZeneca-NGS/disambiguate"
+      tool_dev_url: "https://github.com/AstraZeneca-NGS/disambiguate"
+      doi: "10.12688/f1000research.10082.2"
+      licence:
+        - "MIT"
+      identifier: biotools:disambiguate
+      args_id: "$args"
+input:
+  - - meta:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'sample1' ]`
+    - bam_a:
+        type: file
+        description: BAM file with alignments to species A (e.g. human). Mandatory.
+        pattern: "*.{bam}"
+        ontologies:
+          - edam: "http://edamontology.org/format_2572"
+    - bam_b:
+        type: file
+        description: BAM file with alignments to species B (e.g. mouse). Mandatory.
+        pattern: "*.{bam}"
+        ontologies:
+          - edam: "http://edamontology.org/format_2572"
+output:
+  disambiguated_bam_a:
+    - - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. `[ id:'sample1' ]`
+      - "*disambiguatedSpeciesA.bam":
+          type: file
+          description: Reads uniquely assigned to species A.
+          pattern: "*disambiguatedSpeciesA.bam"
+          ontologies:
+            - edam: "http://edamontology.org/format_2572"
+  disambiguated_bam_b:
+    - - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. `[ id:'sample1' ]`
+      - "*disambiguatedSpeciesB.bam":
+          type: file
+          description: Reads uniquely assigned to species B.
+          pattern: "*disambiguatedSpeciesB.bam"
+          ontologies:
+            - edam: "http://edamontology.org/format_2572"
+  ambiguous_bam_a:
+    - - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. `[ id:'sample1' ]`
+      - "*ambiguousSpeciesA.bam":
+          type: file
+          description: Reads aligned to species A that also aligned to species B but
+            could not be uniquely assigned.
+          pattern: "*ambiguousSpeciesA.bam"
+          ontologies:
+            - edam: "http://edamontology.org/format_2572"
+  ambiguous_bam_b:
+    - - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. `[ id:'sample1' ]`
+      - "*ambiguousSpeciesB.bam":
+          type: file
+          description: Reads aligned to species B that also aligned to species A but
+            could not be uniquely assigned.
+          pattern: "*ambiguousSpeciesB.bam"
+          ontologies:
+            - edam: "http://edamontology.org/format_2572"
+  summary:
+    - - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. `[ id:'sample1' ]`
+      - "*_summary.txt":
+          type: file
+          description: Summary of read counts uniquely assigned to species A, species
+            B, and ambiguous reads.
+          pattern: "*_summary.txt"
+          ontologies:
+            - edam: "http://edamontology.org/format_2330"
+  versions_ngs_disambiguate:
+    - - ${task.process}:
+          type: string
+          description: The name of the process
+      - ngs-disambiguate:
+          type: string
+          description: The name of the tool
+      - 2018.05.03:
+          type: string
+          description: The expression to obtain the version of the tool
+topics:
+  versions:
+    - - ${task.process}:
+          type: string
+          description: The name of the process
+      - ngs-disambiguate:
+          type: string
+          description: The name of the tool
+      - 2018.05.03:
+          type: string
+          description: The expression to obtain the version of the tool
+authors:
+  - "@ctuni"
+maintainers:
+  - "@ctuni"

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

@@ -0,0 +1,59 @@
+nextflow_process {
+
+    name "Test Process DISAMBIGUATE"
+    script "../main.nf"
+    process "DISAMBIGUATE"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "disambiguate"
+
+    test("homo_sapiens - pipistrellus_nathusii - bam") {
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id:'test' ],
+                    file(params.modules_testdata_base_path + 'genomics/homo_sapiens/illumina/bam/test.paired_end.sorted.bam', checkIfExists: true),
+                    file(params.modules_testdata_base_path + 'genomics/eukaryotes/Pipistrellus_nathusii/genome/test.bam', checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(sanitizeOutput(process.out)).match() }
+            )
+        }
+
+    }
+
+    test("homo_sapiens - pipistrellus_nathusii - bam - stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id:'test' ],
+                    file(params.modules_testdata_base_path + 'genomics/homo_sapiens/illumina/bam/test.paired_end.sorted.bam', checkIfExists: true),
+                    file(params.modules_testdata_base_path + 'genomics/eukaryotes/Pipistrellus_nathusii/genome/test.bam', checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(sanitizeOutput(process.out)).match() }
+            )
+        }
+
+    }
+
+}

modules/nf-core/disambiguate/tests/main.nf.test.snap

@@ -0,0 +1,118 @@
+{
+    "homo_sapiens - pipistrellus_nathusii - bam": {
+        "content": [
+            {
+                "ambiguous_bam_a": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.ambiguousSpeciesA.bam:md5,ebd81596ea7369e28f79326972b20af4"
+                    ]
+                ],
+                "ambiguous_bam_b": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.ambiguousSpeciesB.bam:md5,4ba9631d1fbfe3308b5d36eddb76dbe1"
+                    ]
+                ],
+                "disambiguated_bam_a": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.disambiguatedSpeciesA.bam:md5,15105f572c62cdfcc5c4c58d3e1f072c"
+                    ]
+                ],
+                "disambiguated_bam_b": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.disambiguatedSpeciesB.bam:md5,27d054e54e112840da5f9a9de874e652"
+                    ]
+                ],
+                "summary": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_summary.txt:md5,f3a14c362e13c08e1e87cabb2e0b2b78"
+                    ]
+                ],
+                "versions_ngs_disambiguate": [
+                    [
+                        "DISAMBIGUATE",
+                        "ngs-disambiguate",
+                        "2018.05.03"
+                    ]
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.5"
+        },
+        "timestamp": "2026-03-19T10:27:04.503780051"
+    },
+    "homo_sapiens - pipistrellus_nathusii - bam - stub": {
+        "content": [
+            {
+                "ambiguous_bam_a": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.ambiguousSpeciesA.bam:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "ambiguous_bam_b": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.ambiguousSpeciesB.bam:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "disambiguated_bam_a": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.disambiguatedSpeciesA.bam:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "disambiguated_bam_b": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.disambiguatedSpeciesB.bam:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "summary": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_summary.txt:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "versions_ngs_disambiguate": [
+                    [
+                        "DISAMBIGUATE",
+                        "ngs-disambiguate",
+                        "2018.05.03"
+                    ]
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.5"
+        },
+        "timestamp": "2026-03-19T10:27:22.859032297"
+    }
+}