Add UPP (#6546)

* Add UPP * fix lint * Update UPP * fix prettier * Update modules/nf-core/upp/align/tests/main.nf.test Co-authored-by: Simon Pearce <[email protected]> * Update modules/nf-core/upp/align/tests/main.nf.test Co-authored-by: Simon Pearce <[email protected]> * fixes * update * update tests * update * update * Update modules/nf-core/upp/align/main.nf Co-authored-by: Simon Pearce <[email protected]> --------- Co-authored-by: Simon Pearce <[email protected]>
nf-core · Sep 17, 2024 · 3be751e · 3be751e
1 parent 7b9ce4b
commit 3be751e
Show file tree

Hide file tree

Showing 7 changed files with 340 additions and 0 deletions.
diff --git a/modules/nf-core/upp/align/environment.yml b/modules/nf-core/upp/align/environment.yml
@@ -0,0 +1,6 @@
+channels:
+  - conda-forge
+  - bioconda
+dependencies:
+  - bioconda::sepp=4.5.5
+  - conda-forge::pigz=2.8
diff --git a/modules/nf-core/upp/align/main.nf b/modules/nf-core/upp/align/main.nf
@@ -0,0 +1,71 @@
+process UPP_ALIGN {
+    tag "$meta.id"
+    label 'process_medium'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'oras://community.wave.seqera.io/library/sepp_pigz:d72591720d0277b1':
+        'community.wave.seqera.io/library/sepp_pigz:ea6dbc7704a2e251' }"
+
+    input:
+    tuple val(meta) , path(fasta)
+    tuple val(meta2), path(tree)
+    val(compress)
+
+    output:
+    tuple val(meta), path("*.aln{.gz,}"), emit: alignment
+    path "versions.yml"                 , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    def tree_args = tree ? "-t $tree" : ""
+    """
+
+    if [ "$workflow.containerEngine" = 'singularity' ]; then
+        export CONDA_PREFIX="/opt/conda/"
+        export PASTA_TOOLS_DEVDIR="/opt/conda/bin/"
+    fi
+
+    run_upp.py \\
+        $args \\
+        -x $task.cpus \\
+        -s ${fasta} \\
+        -d . \\
+        -o ${prefix} \\
+        -p ./upp-temporary
+
+    mv ${prefix}_alignment.fasta ${prefix}.aln
+
+    if ${compress}; then
+        pigz -p ${task.cpus} ${prefix}.aln
+    fi
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        upp: \$(run_upp.py -v | grep "run_upp" | cut -f2 -d" ")
+        pigz: \$(echo \$(pigz --version 2>&1) | sed 's/^.*pigz\\w*//' ))
+    END_VERSIONS
+    """
+
+    stub:
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+
+
+    if [ "$compress" = true ]; then
+        echo | gzip > "${prefix}.aln.gz"
+    else
+        touch "${prefix}.aln"
+    fi
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        upp: \$(run_upp.py -v | grep "run_upp" | cut -f2 -d" ")
+        pigz: \$(echo \$(pigz --version 2>&1) | sed 's/^.*pigz\\w*//' ))
+    END_VERSIONS
+    """
+}
diff --git a/modules/nf-core/upp/align/meta.yml b/modules/nf-core/upp/align/meta.yml
@@ -0,0 +1,57 @@
+name: "upp_align"
+description: Aligns protein structures using UPP
+keywords:
+  - alignment
+  - MSA
+  - genomics
+  - structure
+tools:
+  - "upp":
+      description: "SATe-enabled phylogenetic placement"
+      homepage: "https://github.com/smirarab/sepp/tree/master"
+      documentation: "https://github.com/smirarab/sepp/blob/master/README.UPP.md"
+      tool_dev_url: "https://github.com/smirarab/sepp/tree/master"
+      doi: "10.1093/bioinformatics/btad007"
+      licence: ["GPL v3"]
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. `[ id:'test']`
+  - fasta:
+      type: file
+      description: Input sequences in FASTA format
+      pattern: "*.{fa,fasta}"
+  - meta2:
+      type: map
+      description: |
+        Groovy Map containing tree information
+        e.g. `[ id:'test_tree']`
+  - tree:
+      type: file
+      description: Input guide tree in Newick format
+      pattern: "*.{dnd}"
+  - compress:
+      type: boolean
+      description: Flag representing whether the output MSA should be compressed. Set to true to enable/false to disable compression.
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. `[ id:'test']`
+  - alignment:
+      type: file
+      description: Alignment file, in FASTA format. May be gzipped or uncompressed, depending on if compress is set to true or false
+      pattern: "*.aln{.gz,}"
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+
+authors:
+  - "@luisas"
+maintainers:
+  - "@luisas"
diff --git a/modules/nf-core/upp/align/tests/main.nf.test b/modules/nf-core/upp/align/tests/main.nf.test
@@ -0,0 +1,96 @@
+nextflow_process {
+
+    name "Test Process UPP_ALIGN"
+    script "../main.nf"
+    process "UPP_ALIGN"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "upp"
+    tag "upp/align"
+    tag "famsa/guidetree"
+
+    test("fasta - align_sequence - uncompressed") {
+        config "./nextflow.config"
+
+        when {
+            process {
+                """
+                    input[0] = [ [ id:'test' ],
+                    file("https://raw.githubusercontent.com/nf-core/test-datasets/multiplesequencealign/testdata/setoxin-ref.fa", checkIfExists: true)
+                    ]
+                    input[1] = [[:],[]]
+                    input[2] = false
+                    """
+            }
+        }
+
+        then {
+            assertAll(
+                    { assert process.success },
+                    { assert snapshot(process.out).match() }
+                    )
+        }
+    }
+
+    test("fasta - with_tree - compressed") {
+        config "./nextflow.config"
+
+        setup {
+
+            run("FAMSA_GUIDETREE") {
+                script "../../../famsa/guidetree/main.nf"
+                process {
+                    """
+                    input[0] = [ [ id:'tree' ],
+                                file("https://raw.githubusercontent.com/nf-core/test-datasets/multiplesequencealign/testdata/setoxin-ref.fa", checkIfExists: true)
+                               ]
+
+                    """
+                }
+            }
+        }
+        when {
+            process {
+                """
+                    input[0] = [ [ id:'test_tree' ],
+                    file("https://raw.githubusercontent.com/nf-core/test-datasets/multiplesequencealign/testdata/setoxin-ref.fa", checkIfExists: true)
+                    ]
+                    input[1] = FAMSA_GUIDETREE.out.tree.collect{ meta, tree -> tree }.map{ tree -> [[ id: 'test_tree'], tree]}
+                    input[2] = true
+                    """
+            }
+        }
+
+        then {
+            assertAll(
+                    { assert process.success },
+                    { assert snapshot(process.out).match()}
+            )
+        }
+    }
+
+    test("stub") {
+        config "./nextflow.config"
+
+        options "-stub"
+        when {
+            process {
+                """
+                    input[0] = [ [ id:'test' ],
+                    file("https://raw.githubusercontent.com/nf-core/test-datasets/multiplesequencealign/testdata/setoxin-ref.fa", checkIfExists: true)
+                    ]
+                    input[1] = [[:],[]]
+                    input[2] = false
+                    """
+            }
+        }
+
+        then {
+            assertAll(
+                    { assert process.success },
+                    { assert snapshot(process.out).match()}
+                    )
+        }
+    }
+}
diff --git a/modules/nf-core/upp/align/tests/main.nf.test.snap b/modules/nf-core/upp/align/tests/main.nf.test.snap
@@ -0,0 +1,101 @@
+{
+    "fasta - align_sequence - uncompressed": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.aln:md5,e6b5291e9cdb40e9b7c72688e4da533b"
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,b431bb15ae86dcd4485d921df1752a98"
+                ],
+                "alignment": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.aln:md5,e6b5291e9cdb40e9b7c72688e4da533b"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,b431bb15ae86dcd4485d921df1752a98"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.0",
+            "nextflow": "24.04.4"
+        },
+        "timestamp": "2024-09-17T07:51:30.876772941"
+    },
+    "stub": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.aln:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,b431bb15ae86dcd4485d921df1752a98"
+                ],
+                "alignment": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.aln:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,b431bb15ae86dcd4485d921df1752a98"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.0",
+            "nextflow": "24.04.4"
+        },
+        "timestamp": "2024-09-17T07:36:19.135281969"
+    },
+    "fasta - with_tree - compressed": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test_tree"
+                        },
+                        "test_tree.aln.gz:md5,e6b5291e9cdb40e9b7c72688e4da533b"
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,b431bb15ae86dcd4485d921df1752a98"
+                ],
+                "alignment": [
+                    [
+                        {
+                            "id": "test_tree"
+                        },
+                        "test_tree.aln.gz:md5,e6b5291e9cdb40e9b7c72688e4da533b"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,b431bb15ae86dcd4485d921df1752a98"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.0",
+            "nextflow": "24.04.4"
+        },
+        "timestamp": "2024-09-17T07:52:05.47226891"
+    }
+}
diff --git a/modules/nf-core/upp/align/tests/nextflow.config b/modules/nf-core/upp/align/tests/nextflow.config
@@ -0,0 +1,7 @@
+process {
+
+    withName: "UPP_ALIGN" {
+        ext.args = { "-m amino" }
+    }
+
+}
diff --git a/modules/nf-core/upp/align/tests/tags.yml b/modules/nf-core/upp/align/tests/tags.yml
@@ -0,0 +1,2 @@
+upp/align:
+  - "modules/nf-core/upp/align/**"