Foldmason (nf-core#6355)

* Add foldmason * Add test * fix prettier * fix prettier * Update modules/nf-core/foldmason/easymsa/tests/main.nf.test Co-authored-by: Simon Pearce <[email protected]> * Update modules/nf-core/foldmason/easymsa/main.nf Co-authored-by: Simon Pearce <[email protected]> * Update modules/nf-core/foldmason/easymsa/tests/main.nf.test Co-authored-by: Simon Pearce <[email protected]> * Update main.nf * Update main.nf.test * Update meta.yml * Update modules/nf-core/foldmason/easymsa/tests/main.nf.test Co-authored-by: Simon Pearce <[email protected]> * Update modules/nf-core/foldmason/easymsa/tests/main.nf.test Co-authored-by: Simon Pearce <[email protected]> * Add stub run tests --------- Co-authored-by: Simon Pearce <[email protected]>
famosab · Aug 28, 2024 · 5a38b71 · 5a38b71
1 parent a35bf30
commit 5a38b71
Show file tree

Hide file tree

Showing 6 changed files with 385 additions and 0 deletions.
diff --git a/modules/nf-core/foldmason/easymsa/environment.yml b/modules/nf-core/foldmason/easymsa/environment.yml
@@ -0,0 +1,10 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/environment-schema.json
+name: "foldmason_easymsa"
+channels:
+  - conda-forge
+  - bioconda
+  - defaults
+dependencies:
+  - bioconda::foldmason=1.763a428
+  - conda-forge::pigz=2.8
diff --git a/modules/nf-core/foldmason/easymsa/main.nf b/modules/nf-core/foldmason/easymsa/main.nf
@@ -0,0 +1,55 @@
+process FOLDMASON_EASYMSA {
+    tag "$meta.id"
+    label 'process_medium'
+
+    conda "${moduleDir}/environment.yml"
+    container "community.wave.seqera.io/library/foldmason_pigz:97b3311addb0f4a7"
+
+    input:
+    tuple val(meta), path(pdbs)
+    val(compress)
+
+    output:
+    tuple val(meta), path("${prefix}_3di.fa${compress ? '.gz' : ''}"), emit: msa_3di
+    tuple val(meta), path("${prefix}_aa.fa${compress ? '.gz' : ''}"), emit: msa_aa
+    path "versions.yml"           , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    foldmason easy-msa \\
+        $args \\
+        --threads $task.cpus \\
+        ${pdbs} \\
+        ${prefix} \\
+        tmp
+
+    if ${compress}; then
+        pigz -p ${task.cpus} ${prefix}.fa
+    fi
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        foldmason: \$(foldmason | grep "foldmason Version:" | cut -d":" -f 2 | awk '{\$1=\$1;print}')
+        pigz: \$(echo \$(pigz --version 2>&1) | sed 's/^.*pigz\\w*//' ))
+    END_VERSIONS
+    """
+
+    stub:
+    def args = task.ext.args ?: ''
+    prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    echo ""  ${compress ? '| gzip' : ''} > ${prefix}_3di.fa${compress ? '.gz' : ''}
+    echo ""  ${compress ? '| gzip' : ''} > ${prefix}_aa.fa${compress ? '.gz' : ''}
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        foldmason: \$(foldmason | grep "foldmason Version:" | cut -d":" -f 2 | awk '{\$1=\$1;print}')
+        pigz: \$(echo \$(pigz --version 2>&1) | sed 's/^.*pigz\\w*//' ))
+    END_VERSIONS
+    """
+}
diff --git a/modules/nf-core/foldmason/easymsa/meta.yml b/modules/nf-core/foldmason/easymsa/meta.yml
@@ -0,0 +1,55 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/meta-schema.json
+name: "foldmason_easymsa"
+description: Aligns protein structures using foldmason
+keywords:
+  - alignment
+  - MSA
+  - genomics
+  - structure
+tools:
+  - "foldmason":
+      description: "Multiple Protein Structure Alignment at Scale with FoldMason"
+      homepage: "https://github.com/steineggerlab/foldmason"
+      documentation: "https://github.com/steineggerlab/foldmason"
+      tool_dev_url: "https://github.com/steineggerlab/foldmason"
+      doi: "10.1101/2024.08.01.606130"
+      licence: ["GPL v3"]
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. `[ id:'sample1', single_end:false ]`
+  - pdbs:
+      type: file
+      description: Input protein structures in PDB format.
+      pattern: "*.{pdb,mmcif}"
+  - compress:
+      type: boolean
+      description: Flag representing whether the output MSA should be compressed. Set to true to enable/false to disable compression. Compression is done using pigz, and is multithreaded.
+
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. `[ id:'sample1', single_end:false ]`
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+  - msa_3di:
+      type: file
+      description: Fasta file containing the multiple sequence alignment with 3Di alphabet
+      pattern: "*.{fa}"
+  - msa_aa:
+      type: file
+      description: Fasta file containing the multiple sequence alignment with Amino Acid alphabet
+      pattern: "*.{fa}"
+
+authors:
+  - "@luisas"
+maintainers:
+  - "@luisas"
diff --git a/modules/nf-core/foldmason/easymsa/tests/main.nf.test b/modules/nf-core/foldmason/easymsa/tests/main.nf.test
@@ -0,0 +1,114 @@
+nextflow_process {
+
+    name "Test Process FOLDMASON_EASYMSA"
+    script "../main.nf"
+    process "FOLDMASON_EASYMSA"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "foldmason"
+    tag "foldmason/easymsa"
+    tag "untar"
+
+    test("Test on seatoxin dataset - uncompressed") {
+        setup {
+
+            run("UNTAR") {
+                script "../../../../../modules/nf-core/untar/main.nf"
+                    process {
+                        """
+                            archive = file("https://raw.githubusercontent.com/nf-core/test-datasets/multiplesequencealign/testdata/structures/seatoxin-ref.tar.gz", checkIfExists: true)
+                            input[0] = Channel.of(tuple([id:'test'], archive))
+                        """
+                    }
+            }
+        }
+
+        when {
+            params {
+            }
+            process {
+                """
+                    input[0] = UNTAR.out.untar.map { meta,dir -> [meta, file(dir).listFiles().collect()]}
+                    input[1] = false
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success  },
+                { assert snapshot(process.out).match() }
+                )
+        }
+    }
+
+    test("Test on seatoxin dataset - compressed") {
+        setup {
+
+            run("UNTAR") {
+                script "../../../../../modules/nf-core/untar/main.nf"
+                    process {
+                        """
+                            archive = file("https://raw.githubusercontent.com/nf-core/test-datasets/multiplesequencealign/testdata/structures/seatoxin-ref.tar.gz", checkIfExists: true)
+                            input[0] = Channel.of(tuple([id:'test'], archive))
+                        """
+                    }
+            }
+        }
+
+        when {
+            params {
+            }
+            process {
+                """
+                    input[0] = UNTAR.out.untar.map { meta,dir -> [[ id:'test' ], file(dir).listFiles().collect()]}
+                    input[1] = false
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success  },
+                { assert snapshot(process.out).match() }
+                )
+        }
+    }
+
+    test("Stub run") {
+        options "-stub"
+
+        setup {
+
+            run("UNTAR") {
+                script "../../../../../modules/nf-core/untar/main.nf"
+                    process {
+                        """
+                            archive = file("https://raw.githubusercontent.com/nf-core/test-datasets/multiplesequencealign/testdata/structures/seatoxin-ref.tar.gz", checkIfExists: true)
+                            input[0] = Channel.of(tuple([id:'test'], archive))
+                        """
+                    }
+            }
+        }
+
+        when {
+            params {
+            }
+            process {
+                """
+                    input[0] = UNTAR.out.untar.map { meta,dir -> [meta, file(dir).listFiles().collect()]}
+                    input[1] = false
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success  },
+                { assert snapshot(process.out).match() }
+                )
+        }
+    }
+
+}
diff --git a/modules/nf-core/foldmason/easymsa/tests/main.nf.test.snap b/modules/nf-core/foldmason/easymsa/tests/main.nf.test.snap
@@ -0,0 +1,149 @@
+{
+    "Stub run": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_3di.fa:md5,68b329da9893e34099c7d8ad5cb9c940"
+                    ]
+                ],
+                "1": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_aa.fa:md5,68b329da9893e34099c7d8ad5cb9c940"
+                    ]
+                ],
+                "2": [
+                    "versions.yml:md5,da4694171d1b0bb9559f7049334126ed"
+                ],
+                "msa_3di": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_3di.fa:md5,68b329da9893e34099c7d8ad5cb9c940"
+                    ]
+                ],
+                "msa_aa": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_aa.fa:md5,68b329da9893e34099c7d8ad5cb9c940"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,da4694171d1b0bb9559f7049334126ed"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.0",
+            "nextflow": "24.04.4"
+        },
+        "timestamp": "2024-08-28T09:10:12.591561643"
+    },
+    "Test on seatoxin dataset - uncompressed": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_3di.fa:md5,ba1b6c8c5df11bdebfea12070bd9cb17"
+                    ]
+                ],
+                "1": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_aa.fa:md5,33e93479603115b46ef76af3f6a20cf1"
+                    ]
+                ],
+                "2": [
+                    "versions.yml:md5,da4694171d1b0bb9559f7049334126ed"
+                ],
+                "msa_3di": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_3di.fa:md5,ba1b6c8c5df11bdebfea12070bd9cb17"
+                    ]
+                ],
+                "msa_aa": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_aa.fa:md5,33e93479603115b46ef76af3f6a20cf1"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,da4694171d1b0bb9559f7049334126ed"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.0",
+            "nextflow": "24.04.4"
+        },
+        "timestamp": "2024-08-28T08:58:09.52040475"
+    },
+    "Test on seatoxin dataset - compressed": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_3di.fa:md5,ba1b6c8c5df11bdebfea12070bd9cb17"
+                    ]
+                ],
+                "1": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_aa.fa:md5,33e93479603115b46ef76af3f6a20cf1"
+                    ]
+                ],
+                "2": [
+                    "versions.yml:md5,da4694171d1b0bb9559f7049334126ed"
+                ],
+                "msa_3di": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_3di.fa:md5,ba1b6c8c5df11bdebfea12070bd9cb17"
+                    ]
+                ],
+                "msa_aa": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test_aa.fa:md5,33e93479603115b46ef76af3f6a20cf1"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,da4694171d1b0bb9559f7049334126ed"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.0",
+            "nextflow": "24.04.4"
+        },
+        "timestamp": "2024-08-28T08:58:21.721568566"
+    }
+}
diff --git a/modules/nf-core/foldmason/easymsa/tests/tags.yml b/modules/nf-core/foldmason/easymsa/tests/tags.yml
@@ -0,0 +1,2 @@
+foldmason/easymsa:
+  - "modules/nf-core/foldmason/easymsa/**"