From bdca0f17ea92b6290e458bbcef71544041d7b720 Mon Sep 17 00:00:00 2001
From: Dave Carlson <david.carlson@stonybrook.edu>
Date: Mon, 13 May 2024 23:55:19 +0800
Subject: [PATCH] add bedops/gtf2bed (#4476)

* add gt2bed main.nf

* add meta.yml file

* bug fixes and add tests

* run prettier on yml files

* replace meta.yml

* prettier formatting

* Update modules/nf-core/bedops/gtf2bed/main.nf

Co-authored-by: Maxime U Garcia <maxime.garcia@seqera.io>

* Swap to nftest

* Update snap

---------

Co-authored-by: Maxime U Garcia <max.u.garcia@gmail.com>
Co-authored-by: Maxime U Garcia <maxime.garcia@seqera.io>
Co-authored-by: Simon Pearce <24893913+SPPearce@users.noreply.github.com>
---
 .../nf-core/bedops/gtf2bed/environment.yml    |  7 ++
 modules/nf-core/bedops/gtf2bed/main.nf        | 49 +++++++++++++
 modules/nf-core/bedops/gtf2bed/meta.yml       | 38 +++++++++++
 .../nf-core/bedops/gtf2bed/tests/main.nf.test | 56 +++++++++++++++
 .../bedops/gtf2bed/tests/main.nf.test.snap    | 68 +++++++++++++++++++
 modules/nf-core/bedops/gtf2bed/tests/tags.yml |  2 +
 6 files changed, 220 insertions(+)
 create mode 100644 modules/nf-core/bedops/gtf2bed/environment.yml
 create mode 100644 modules/nf-core/bedops/gtf2bed/main.nf
 create mode 100644 modules/nf-core/bedops/gtf2bed/meta.yml
 create mode 100644 modules/nf-core/bedops/gtf2bed/tests/main.nf.test
 create mode 100644 modules/nf-core/bedops/gtf2bed/tests/main.nf.test.snap
 create mode 100644 modules/nf-core/bedops/gtf2bed/tests/tags.yml

diff --git a/modules/nf-core/bedops/gtf2bed/environment.yml b/modules/nf-core/bedops/gtf2bed/environment.yml
new file mode 100644
index 00000000000..8aede242838
--- /dev/null
+++ b/modules/nf-core/bedops/gtf2bed/environment.yml
@@ -0,0 +1,7 @@
+name: "bedops_gtf2bed"
+channels:
+  - conda-forge
+  - bioconda
+  - defaults
+dependencies:
+  - "bioconda::bedops=2.4.41"
diff --git a/modules/nf-core/bedops/gtf2bed/main.nf b/modules/nf-core/bedops/gtf2bed/main.nf
new file mode 100644
index 00000000000..13a37b5cc85
--- /dev/null
+++ b/modules/nf-core/bedops/gtf2bed/main.nf
@@ -0,0 +1,49 @@
+process BEDOPS_GTF2BED {
+    tag "$gtf"
+    label 'process_low'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/bedops:2.4.41--h4ac6f70_2':
+        'biocontainers/bedops:2.4.41--h4ac6f70_2' }"
+
+    input:
+    tuple val(meta), path(gtf)
+
+    output:
+    tuple val(meta), path('*.bed'), emit: bed
+    path "versions.yml"           , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args   = task.ext.args   ?: ''
+    def prefix = task.ext.prefix ?: "${gtf.baseName}"
+
+    """
+    cat \\
+    $gtf \\
+    | gtf2bed \\
+    $args \\
+    --attribute-key=exon_id \\
+    > ${prefix}.bed
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        gtf2bed: \$(bedops --version | grep version | awk ' { print \$2 } ')
+    END_VERSIONS
+    """
+
+    stub:
+    def prefix = task.ext.prefix ?: "${gtf.baseName}"
+    """
+    touch ${prefix}.bed
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        gtf2bed: \$(bedops --version | grep version | awk ' { print \$2 } ')
+    END_VERSIONS
+    """
+
+}
diff --git a/modules/nf-core/bedops/gtf2bed/meta.yml b/modules/nf-core/bedops/gtf2bed/meta.yml
new file mode 100644
index 00000000000..2646b74a2f8
--- /dev/null
+++ b/modules/nf-core/bedops/gtf2bed/meta.yml
@@ -0,0 +1,38 @@
+name: "bedops_gtf2bed"
+description: Convert gtf format to bed format
+keywords:
+  - gtf
+  - bed
+  - conversion
+tools:
+  - gtf2bed:
+      description: The gtf2bed script converts 1-based, closed [start, end] Gene Transfer Format v2.2 (GTF2.2) to sorted, 0-based, half-open [start-1, end) extended BED-formatted data.
+      homepage: https://bedops.readthedocs.io/en/latest/content/reference/file-management/conversion/gtf2bed.html
+      documentation: https://bedops.readthedocs.io/en/latest/content/reference/file-management/conversion/gtf2bed.html
+      tool_dev_url: https://github.com/bedops/bedops
+      doi: 10.1093/bioinformatics/bts277
+      licence: ["GPL v2"]
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. `[ id:'sample1', single_end:false ]`
+
+  - gtf:
+      type: file
+      description: A reference file in GTF format
+      pattern: "*.{gtf,gtf.gz}"
+
+output:
+  - bed:
+      type: file
+      description: A reference file in BED format
+      pattern: "*.{bed}"
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+authors:
+  - "@davidecarlson"
diff --git a/modules/nf-core/bedops/gtf2bed/tests/main.nf.test b/modules/nf-core/bedops/gtf2bed/tests/main.nf.test
new file mode 100644
index 00000000000..8dcc67f5760
--- /dev/null
+++ b/modules/nf-core/bedops/gtf2bed/tests/main.nf.test
@@ -0,0 +1,56 @@
+nextflow_process {
+
+    name "Test Process BEDOPS_GTF2BED"
+    script "../main.nf"
+    process "BEDOPS_GTF2BED"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "bedops"
+    tag "bedops/gtf2bed"
+
+    test("homo sapiens") {
+
+        when {
+            process {
+                """
+                input[0] = [[ id:'test' ], // meta map
+                             file(params.test_data['homo_sapiens']['genome']['genome_gtf'], checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+    test("homo sapiens - stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+                input[0] = [[ id:'test' ], // meta map
+                             file(params.test_data['homo_sapiens']['genome']['genome_gtf'], checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+}
diff --git a/modules/nf-core/bedops/gtf2bed/tests/main.nf.test.snap b/modules/nf-core/bedops/gtf2bed/tests/main.nf.test.snap
new file mode 100644
index 00000000000..9c9f905e43c
--- /dev/null
+++ b/modules/nf-core/bedops/gtf2bed/tests/main.nf.test.snap
@@ -0,0 +1,68 @@
+{
+    "homo sapiens": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "genome.bed:md5,783257789b0c59a1a5391f7f2b9ac08a"
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,fb301f89b7f040eedb8ea7edaaf05187"
+                ],
+                "bed": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "genome.bed:md5,783257789b0c59a1a5391f7f2b9ac08a"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,fb301f89b7f040eedb8ea7edaaf05187"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.8.4",
+            "nextflow": "23.10.1"
+        },
+        "timestamp": "2024-05-13T15:45:25.277799374"
+    },
+    "homo sapiens - stub": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "genome.bed:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,fb301f89b7f040eedb8ea7edaaf05187"
+                ],
+                "bed": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "genome.bed:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,fb301f89b7f040eedb8ea7edaaf05187"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.8.4",
+            "nextflow": "23.10.1"
+        },
+        "timestamp": "2024-05-13T15:45:34.937117692"
+    }
+}
\ No newline at end of file
diff --git a/modules/nf-core/bedops/gtf2bed/tests/tags.yml b/modules/nf-core/bedops/gtf2bed/tests/tags.yml
new file mode 100644
index 00000000000..215ef9a1711
--- /dev/null
+++ b/modules/nf-core/bedops/gtf2bed/tests/tags.yml
@@ -0,0 +1,2 @@
+bedops/gtf2bed:
+  - "modules/nf-core/bedops/gtf2bed/**"