initial commit of kma_index (nf-core#7236)

* initial commit of kma_index * trying to fix linting for meta.yml * fix linting of meta.yml again * fixing DOI for linting * fixing file type for linting * fix output file name for linting * add snapshot * update output format * upgrade latest container build trying to satisfy docker CI
atrigila · Dec 20, 2024 · 3d107f4 · 3d107f4
1 parent cc217b9
commit 3d107f4
Show file tree

Hide file tree

Showing 6 changed files with 264 additions and 0 deletions.
diff --git a/modules/nf-core/kma/index/environment.yml b/modules/nf-core/kma/index/environment.yml
@@ -0,0 +1,7 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/environment-schema.json
+channels:
+  - bioconda
+  - conda-forge
+dependencies:
+  - bioconda::kma=1.4.15
diff --git a/modules/nf-core/kma/index/main.nf b/modules/nf-core/kma/index/main.nf
@@ -0,0 +1,52 @@
+process KMA_INDEX {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/kma:1.4.15--h577a1d6_1' :
+        'biocontainers/kma:1.4.15--h577a1d6_1' }"
+
+    input:
+    tuple val(meta), path(fasta)
+
+    output:
+    tuple val(meta), path("kmaindex"),  emit: index
+    path "versions.yml",                emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def prefix  = task.ext.prefix ?: "${fasta.baseName}"
+    def args    = task.ext.args ?: ''
+    """
+    mkdir kmaindex
+    kma \\
+        index \\
+        -i ${fasta} \\
+        -o kmaindex/${prefix} \\
+        $args
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        kma: \$(echo \$(kma_index -v 2>&1) | sed 's/^KMA_index-\$//')
+    END_VERSIONS
+    """
+
+    stub:
+    def prefix  = task.ext.prefix ?: "${fasta.baseName}"
+    """
+    mkdir kmaindex
+
+    touch kmaindex/${prefix}.comp.b
+    touch kmaindex/${prefix}.length.b
+    touch kmaindex/${prefix}.name
+    touch kmaindex/${prefix}.seq.b
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        kma: \$(echo \$(kma_index -v 2>&1) | sed 's/^KMA_index-\$//')
+    END_VERSIONS
+    """
+}
diff --git a/modules/nf-core/kma/index/meta.yml b/modules/nf-core/kma/index/meta.yml
@@ -0,0 +1,52 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/meta-schema.json
+name: "kma_index"
+description: This module wraps the index module of the KMA alignment tool.
+keywords:
+  - alignment
+  - kma
+  - index
+  - database
+  - reads
+tools:
+  - "kma":
+      description: "Rapid and precise alignment of raw reads against redundant databases with KMA"
+      homepage: "https://bitbucket.org/genomicepidemiology/kma/src/master/"
+      documentation: "https://bitbucket.org/genomicepidemiology/kma/src/master/"
+      tool_dev_url: "https://bitbucket.org/genomicepidemiology/kma/src/master/"
+      doi: "10.1186/s12859-018-2336-6"
+      licence:
+        ["http://www.apache.org/licenses/LICENSE-2.0"]
+
+input:
+  - - meta:
+        type: map
+        description: |
+          Groovy Map containing reference information
+          e.g. `[ id:'reference' ]`
+    - fasta:
+        type: file
+        description: (Multi-)FASTA file of your database sequences.
+        pattern: "*.{fa,fasta}"
+
+output:
+  - index:
+    - meta:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'reference' ]`
+    - 'kmaindex':
+        type: directory
+        description: Directory of KMA index files
+        pattern: "*.{comp.b,length.b,name,seq.b}"
+  - versions:
+    - versions.yml:
+        type: file
+        description: File containing software versions
+        pattern: "versions.yml"
+
+authors:
+  - "@krannich479"
+maintainers:
+  - "@krannich479"
diff --git a/modules/nf-core/kma/index/tests/main.nf.test b/modules/nf-core/kma/index/tests/main.nf.test
@@ -0,0 +1,59 @@
+
+nextflow_process {
+
+    name "Test Process KMA_INDEX"
+    script "../main.nf"
+    process "KMA_INDEX"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "kma"
+    tag "kma/index"
+
+    test("sarscov2 - fasta") {
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id:'MT192765.1', single_end:false ], // meta map
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+    test("sarscov2 - fasta - stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id:'MT192765.1', single_end:false ], // meta map
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+}
diff --git a/modules/nf-core/kma/index/tests/main.nf.test.snap b/modules/nf-core/kma/index/tests/main.nf.test.snap
@@ -0,0 +1,92 @@
+{
+    "sarscov2 - fasta - stub": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "MT192765.1",
+                            "single_end": false
+                        },
+                        [
+                            "genome.comp.b:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "genome.length.b:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "genome.name:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "genome.seq.b:md5,d41d8cd98f00b204e9800998ecf8427e"
+                        ]
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,ba9763933bdc811b1e19e466cceae9c8"
+                ],
+                "index": [
+                    [
+                        {
+                            "id": "MT192765.1",
+                            "single_end": false
+                        },
+                        [
+                            "genome.comp.b:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "genome.length.b:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "genome.name:md5,d41d8cd98f00b204e9800998ecf8427e",
+                            "genome.seq.b:md5,d41d8cd98f00b204e9800998ecf8427e"
+                        ]
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,ba9763933bdc811b1e19e466cceae9c8"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.3"
+        },
+        "timestamp": "2024-12-19T17:07:30.875020727"
+    },
+    "sarscov2 - fasta": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "MT192765.1",
+                            "single_end": false
+                        },
+                        [
+                            "genome.comp.b:md5,2d4075cee26606cfb9b3624f8d7fe260",
+                            "genome.length.b:md5,aaded6f5e19db3f3292e7ff9cd0157ec",
+                            "genome.name:md5,e98cac0e614ac65bb458f77245af7d32",
+                            "genome.seq.b:md5,48647a9697263fe218ddc728c2d01641"
+                        ]
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,ba9763933bdc811b1e19e466cceae9c8"
+                ],
+                "index": [
+                    [
+                        {
+                            "id": "MT192765.1",
+                            "single_end": false
+                        },
+                        [
+                            "genome.comp.b:md5,2d4075cee26606cfb9b3624f8d7fe260",
+                            "genome.length.b:md5,aaded6f5e19db3f3292e7ff9cd0157ec",
+                            "genome.name:md5,e98cac0e614ac65bb458f77245af7d32",
+                            "genome.seq.b:md5,48647a9697263fe218ddc728c2d01641"
+                        ]
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,ba9763933bdc811b1e19e466cceae9c8"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.3"
+        },
+        "timestamp": "2024-12-19T17:07:24.8942212"
+    }
+}
diff --git a/modules/nf-core/kma/index/tests/tags.yml b/modules/nf-core/kma/index/tests/tags.yml
@@ -0,0 +1,2 @@
+kma/index:
+  - "modules/nf-core/kma/index/**"