Add new module EAGLE2

modules/nf-core/eagle2/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::eagle2=2.4.1"

modules/nf-core/eagle2/main.nf

@@ -0,0 +1,61 @@
+process EAGLE2 {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/eagle2:2.4.1--h6a68c12_0':
+        'biocontainers/eagle2:2.4.1--h6a68c12_0' }"
+
+    input:
+    tuple val(meta), path(input), path(index), path(ref_vcf), path(ref_index), path(map)
+
+    output:
+    tuple val(meta), path("*.{vcf,vcf.gz,bcf}"), emit: phased_variants
+    path "versions.yml"                        , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    def input_cmd = ref_vcf ? "--vcfTarget $input" : "--vcf $input"
+    def vcfref_cmd = ref_vcf ? "--vcfRef $ref_vcf" : ""
+    def remove_cmd = remove ? "--remove $remove" : ""
+    def exclude_cmd = exclude ? "--exclude $exclude" : ""
+    """
+    eagle \\
+        $input_cmd \\
+        $vcfref_cmd \\
+        --geneticMapFile $map \\
+        $remove_cmd \\
+        $exclude_cmd \\
+        $args \\
+        --numThreads $task.cpus \\
+        --outPrefix ${prefix} \\
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        eagle2: \$(eagle --help | sed -n 's/.*Eagle v\\([0-9.]\\+\\).*/\\1/p')
+    END_VERSIONS
+    """
+
+    stub:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    def extension = args.contains("--vcfOutFormat z") ? "vcf.gz" :
+                    args.contains("--vcfOutFormat v") ? "vcf"    :
+                    args.contains("--vcfOutFormat b") ? "bcf"    :
+                    args.contains("--vcfOutFormat u") ? "bcf"    :
+                    "vcf.gz"
+    def create_cmd = extension.endsWith("gz") ? "echo '' | bgzip >" : "touch"
+    """
+    ${create_cmd} ${prefix}.${extension}
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        eagle2: \$(eagle --help | sed -n 's/.*Eagle v\\([0-9.]\\+\\).*/\\1/p')
+    END_VERSIONS
+    """
+}

modules/nf-core/eagle2/meta.yml

@@ -0,0 +1,80 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/meta-schema.json
+name: "eagle2"
+description: Perform phasing of genotyped cohort
+keywords:
+  - phasing
+  - haplotypes
+  - reference panel
+  - genomics
+tools:
+  - "eagle2":
+      description: "The Eagle software estimates haplotype phase either within a genotyped cohort or using a phased reference panel."
+      homepage: "https://alkesgroup.broadinstitute.org/Eagle/"
+      documentation: "https://alkesgroup.broadinstitute.org/Eagle/"
+      tool_dev_url: "https://github.com/poruloh/Eagle"
+      doi: "10.1038/ng.3679"
+      licence: ['GPL v3-or-later']
+
+input:
+  - - meta:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'sample1', single_end:false ]`
+    - input:
+        type: file
+        description: Vcf file containing genotyped cohort or target samples to phase
+        pattern: "*.{vcf,vcf.gz,bcf}"
+        ontologies:
+          - edam: "http://edamontology.org/format_3016"
+          - edam: "http://edamontology.org/format_3020"
+    - index:
+        type: file
+        description: Input file index
+        pattern: "*.{csi,tbi}"
+        ontologies:
+          - edam: "http://edamontology.org/format_3700"
+    - ref_vcf:
+        type: file
+        description: Reference panel to phase the genotyped cohort
+        pattern: "*.{vcf,vcf.gz,bcf}"
+        ontologies:
+          - edam: "http://edamontology.org/format_3016"
+          - edam: "http://edamontology.org/format_3020"
+    - ref_index:
+        type: file
+        description: Reference panel index
+        pattern: "*.{csi,tbi}"
+        ontologies:
+          - edam: "http://edamontology.org/format_3700"
+    - map:
+        type: file
+        description: Genetic map file
+        pattern: "*.{txt,txt.gz,map,map.gz}"
+
+output:
+  - phased_variants:
+    - meta:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'sample1', single_end:false ]`
+    - "*.{vcf,vcf.gz,bcf}":
+        type: file
+        description: Input phased variants
+        pattern: "*.{vcf,vcf.gz,bcf}"
+        ontologies:
+          - edam: "http://edamontology.org/format_3016"
+          - edam: "http://edamontology.org/format_3020"
+
+  - versions:
+    - "versions.yml":
+        type: file
+        description: File containing software versions
+        pattern: "versions.yml"
+
+authors:
+  - "@LouisLeNezet"
+maintainers:
+  - "@LouisLeNezet"

modules/nf-core/eagle2/tests/main.nf.test

@@ -0,0 +1,104 @@
+nextflow_process {
+
+    name "Test Process EAGLE2"
+    script "../main.nf"
+    process "EAGLE2"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "eagle2"
+
+    config "./nextflow.config"
+
+    test("homo sapiens - vcf, [], map") {
+        when {
+            params {
+                eagle2_args = "--vcfOutFormat z --chrom chr22"
+            }
+            process {
+                """
+                input[0] = Channel.of([
+                    [ id:'input', single_end:false ], // meta map
+                    file(params.modules_testdata_base_path + "genomics/homo_sapiens/popgen/1000GP.chr22.vcf.gz", checkIfExists: true),
+                    file(params.modules_testdata_base_path + "genomics/homo_sapiens/popgen/1000GP.chr22.vcf.gz.csi", checkIfExists: true),
+                    [], [],
+                    file(params.modules_testdata_base_path + "genomics/homo_sapiens/genome/genome.GRCh38.eagle.22.map.gz", checkIfExists: true)
+                ])
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(
+                    process.out.phased_variants.collect{ meta, vcf -> [meta, file(vcf).name]},
+                    process.out.versions
+                ).match() }
+            )
+        }
+    }
+
+    test("homo sapiens - vcf, ref, map") {
+        when {
+            params {
+                eagle2_args = "--chrom chr22"
+            }
+            process {
+                """
+                input[0] = Channel.of([
+                    [ id:'input', single_end:false ], // meta map
+                    file(params.modules_testdata_base_path + "genomics/homo_sapiens/illumina/vcf/NA12878.chr22.1X.vcf.gz", checkIfExists: true),
+                    file(params.modules_testdata_base_path + "genomics/homo_sapiens/illumina/vcf/NA12878.chr22.1X.vcf.gz.csi", checkIfExists: true),
+                    file(params.modules_testdata_base_path + "genomics/homo_sapiens/popgen/1000GP.chr22.vcf.gz", checkIfExists: true),
+                    file(params.modules_testdata_base_path + "genomics/homo_sapiens/popgen/1000GP.chr22.vcf.gz.csi", checkIfExists: true),
+                    file(params.modules_testdata_base_path + "genomics/homo_sapiens/genome/genome.GRCh38.eagle.22.map.gz", checkIfExists: true)
+                ])
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(
+                    process.out.phased_variants.collect{ meta, vcf -> [meta, file(vcf).name]},
+                    process.out.versions
+                ).match() }
+            )
+        }
+    }
+
+    test("homo sapiens - vcf, [], [] - stub") {
+
+        options "-stub"
+
+        when {
+            params {
+                eagle2_args = "--vcfOutFormat b"
+            }
+            process {
+                """
+                input[0] = Channel.of([
+                    [ id:'input', single_end:false ], // meta map
+                    file(params.modules_testdata_base_path + "genomics/homo_sapiens/popgen/1000GP.chr22.vcf.gz", checkIfExists: true),
+                    file(params.modules_testdata_base_path + "genomics/homo_sapiens/popgen/1000GP.chr22.vcf.gz.csi", checkIfExists: true),
+                    [], [], []
+                ])
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(
+                    process.out,
+                    path(process.out.versions.get(0)).yaml
+                ).match() }
+            )
+        }
+
+    }
+
+}

modules/nf-core/eagle2/tests/main.nf.test.snap

@@ -0,0 +1,84 @@
+{
+    "homo sapiens - vcf, [], map": {
+        "content": [
+            [
+                [
+                    {
+                        "id": "input",
+                        "single_end": false
+                    },
+                    "input.vcf.gz"
+                ]
+            ],
+            [
+                "versions.yml:md5,89d3594c73cc7a875c60ea2762042bbb"
+            ]
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.4"
+        },
+        "timestamp": "2025-03-09T19:41:48.000108431"
+    },
+    "homo sapiens - vcf, [], [] - stub": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "input",
+                            "single_end": false
+                        },
+                        "input.bcf:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,89d3594c73cc7a875c60ea2762042bbb"
+                ],
+                "phased_variants": [
+                    [
+                        {
+                            "id": "input",
+                            "single_end": false
+                        },
+                        "input.bcf:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,89d3594c73cc7a875c60ea2762042bbb"
+                ]
+            },
+            {
+                "EAGLE2": {
+                    "eagle2": "2.4.1"
+                }
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.4"
+        },
+        "timestamp": "2025-03-09T18:27:55.049410748"
+    },
+    "homo sapiens - vcf, ref, map": {
+        "content": [
+            [
+                [
+                    {
+                        "id": "input",
+                        "single_end": false
+                    },
+                    "input.vcf.gz"
+                ]
+            ],
+            [
+                "versions.yml:md5,89d3594c73cc7a875c60ea2762042bbb"
+            ]
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.4"
+        },
+        "timestamp": "2025-03-09T19:41:55.432871651"
+    }
+}

modules/nf-core/eagle2/tests/nextflow.config

@@ -0,0 +1,5 @@
+process {
+    withName: "EAGLE2" {
+        ext.args = {"${params.eagle2_args}"}
+    }
+}