Added repeatmodeler/builddatabase (nf-core#5416)

* Added repeatmodeler/builddatabase * Fixed version extraction * Now doing md5 check on stable db files
asp8200 · Apr 25, 2024 · 1cbb555 · 1cbb555
1 parent c971cdb
commit 1cbb555
Show file tree

Hide file tree

Showing 6 changed files with 257 additions and 0 deletions.
diff --git a/modules/nf-core/repeatmodeler/builddatabase/environment.yml b/modules/nf-core/repeatmodeler/builddatabase/environment.yml
@@ -0,0 +1,9 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/environment-schema.json
+name: "repeatmodeler_builddatabase"
+channels:
+  - conda-forge
+  - bioconda
+  - defaults
+dependencies:
+  - "bioconda::repeatmodeler=2.0.5"
diff --git a/modules/nf-core/repeatmodeler/builddatabase/main.nf b/modules/nf-core/repeatmodeler/builddatabase/main.nf
@@ -0,0 +1,50 @@
+process REPEATMODELER_BUILDDATABASE {
+    tag "$meta.id"
+    label 'process_single'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/repeatmodeler:2.0.5--pl5321hdfd78af_0':
+        'biocontainers/repeatmodeler:2.0.5--pl5321hdfd78af_0' }"
+
+    input:
+    tuple val(meta), path(fasta)
+
+    output:
+    tuple val(meta), path("${prefix}.*")    , emit: db
+    path "versions.yml"                     , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    BuildDatabase \\
+        -name $prefix \\
+        $fasta
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        repeatmodeler: \$(RepeatModeler --version | sed 's/RepeatModeler version //')
+    END_VERSIONS
+    """
+
+    stub:
+    prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    touch ${prefix}.nhr
+    touch ${prefix}.nin
+    touch ${prefix}.njs
+    touch ${prefix}.nnd
+    touch ${prefix}.nni
+    touch ${prefix}.nog
+    touch ${prefix}.nsq
+    touch ${prefix}.translation
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        repeatmodeler: \$(RepeatModeler --version | sed 's/RepeatModeler version //')
+    END_VERSIONS
+    """
+}
diff --git a/modules/nf-core/repeatmodeler/builddatabase/meta.yml b/modules/nf-core/repeatmodeler/builddatabase/meta.yml
@@ -0,0 +1,44 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/meta-schema.json
+name: "repeatmodeler_builddatabase"
+description: Create a database for RepeatModeler
+keywords:
+  - genomics
+  - fasta
+  - repeat
+tools:
+  - "repeatmodeler":
+      description: "RepeatModeler is a de-novo repeat family identification and modeling package."
+      homepage: "https://github.com/Dfam-consortium/RepeatModeler"
+      documentation: "https://github.com/Dfam-consortium/RepeatModeler"
+      tool_dev_url: "https://github.com/Dfam-consortium/RepeatModeler"
+      licence: ["Open Software License v2.1"]
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. `[ id:'sample1' ]`
+  - fasta:
+      type: file
+      description: Fasta file
+      pattern: "*.{fasta,fsa,fa}"
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. `[ id:'sample1' ]`
+  - db:
+      type: file
+      description: Database files for repeatmodeler
+      pattern: "`${prefix}.*`"
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+authors:
+  - "@GallVp"
+maintainers:
+  - "@GallVp"
diff --git a/modules/nf-core/repeatmodeler/builddatabase/tests/main.nf.test b/modules/nf-core/repeatmodeler/builddatabase/tests/main.nf.test
@@ -0,0 +1,60 @@
+nextflow_process {
+
+    name "Test Process REPEATMODELER_BUILDDATABASE"
+    script "../main.nf"
+    process "REPEATMODELER_BUILDDATABASE"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "repeatmodeler"
+    tag "repeatmodeler/builddatabase"
+
+    test("sarscov2-genome_fasta") {
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id:'test' ], // meta map
+                    file(params.test_data['sarscov2']['genome']['genome_fasta'], checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out.versions).match("versions") },
+                { assert snapshot(process.out.db[0][1].collect { file(it).name }.sort().toString()).match("db") },
+                { assert snapshot(process.out.db[0][1].findAll { ! ( "$it"[-3..-1] in [ 'nin', 'njs' ] ) }  ).match("stable_md5") }
+            )
+        }
+
+    }
+
+    test("sarscov2-genome_fasta-stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id:'test' ], // meta map
+                    file(params.test_data['sarscov2']['genome']['genome_fasta'], checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+}
diff --git a/modules/nf-core/repeatmodeler/builddatabase/tests/main.nf.test.snap b/modules/nf-core/repeatmodeler/builddatabase/tests/main.nf.test.snap
@@ -0,0 +1,92 @@
+{
+    "sarscov2-genome_fasta-stub": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        [
+                            "test.nhr:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.nin:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.njs:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.nnd:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.nni:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.nog:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.nsq:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.translation:md5,d41d8cd98f00b204e9800998ecf8427e"
+                        ]
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,7944637266bc3e2726899eaad5e46c87"
+                ],
+                "db": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        [
+                            "test.nhr:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.nin:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.njs:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.nnd:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.nni:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.nog:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.nsq:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "test.translation:md5,d41d8cd98f00b204e9800998ecf8427e"
+                        ]
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,7944637266bc3e2726899eaad5e46c87"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.8.4",
+            "nextflow": "23.10.1"
+        },
+        "timestamp": "2024-04-02T12:06:44.261566"
+    },
+    "versions": {
+        "content": [
+            [
+                "versions.yml:md5,7944637266bc3e2726899eaad5e46c87"
+            ]
+        ],
+        "meta": {
+            "nf-test": "0.8.4",
+            "nextflow": "23.10.1"
+        },
+        "timestamp": "2024-01-09T15:14:48.807063"
+    },
+    "stable_md5": {
+        "content": [
+            [
+                "test.nhr:md5,1a41cb6d0b00c28f62ad60e75ae2f6fc",
+                "test.nnd:md5,2002e13acf59079a1a5782c918894579",
+                "test.nni:md5,26a954ba0fd80983b550d8f6b8b35ff8",
+                "test.nog:md5,30896f123998e926ea2237b89091e7fe",
+                "test.nsq:md5,982cbc7d9e38743b9b1037588862b9da",
+                "test.translation:md5,ccbb119522c09daa976a9015ba999329"
+            ]
+        ],
+        "meta": {
+            "nf-test": "0.8.4",
+            "nextflow": "23.10.1"
+        },
+        "timestamp": "2024-04-23T10:03:41.669433"
+    },
+    "db": {
+        "content": [
+            "[test.nhr, test.nin, test.njs, test.nnd, test.nni, test.nog, test.nsq, test.translation]"
+        ],
+        "meta": {
+            "nf-test": "0.8.4",
+            "nextflow": "23.10.1"
+        },
+        "timestamp": "2024-04-02T12:08:36.94713"
+    }
+}
diff --git a/modules/nf-core/repeatmodeler/builddatabase/tests/tags.yml b/modules/nf-core/repeatmodeler/builddatabase/tests/tags.yml
@@ -0,0 +1,2 @@
+repeatmodeler/builddatabase:
+  - "modules/nf-core/repeatmodeler/builddatabase/**"