From 6342f5b0996c1debb23037efdd1ad7029e801a7f Mon Sep 17 00:00:00 2001
From: Nicolas Vannieuwkerke <nicolas.vannieuwkerke@ugent.be>
Date: Tue, 17 May 2022 14:24:32 +0200
Subject: [PATCH] added bedtools split

---
 modules/bedtools/split/main.nf               | 38 ++++++++++++++++++
 modules/bedtools/split/meta.yml              | 41 ++++++++++++++++++++
 tests/config/pytest_modules.yml              |  4 ++
 tests/modules/bedtools/split/main.nf         | 17 ++++++++
 tests/modules/bedtools/split/nextflow.config |  5 +++
 tests/modules/bedtools/split/test.yml        | 10 +++++
 6 files changed, 115 insertions(+)
 create mode 100644 modules/bedtools/split/main.nf
 create mode 100644 modules/bedtools/split/meta.yml
 create mode 100644 tests/modules/bedtools/split/main.nf
 create mode 100644 tests/modules/bedtools/split/nextflow.config
 create mode 100644 tests/modules/bedtools/split/test.yml

diff --git a/modules/bedtools/split/main.nf b/modules/bedtools/split/main.nf
new file mode 100644
index 00000000..86054292
--- /dev/null
+++ b/modules/bedtools/split/main.nf
@@ -0,0 +1,38 @@
+process BEDTOOLS_SPLIT {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda (params.enable_conda ? "bioconda::bedtools=2.30.0" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/bedtools:2.30.0--h468198e_3':
+        'quay.io/biocontainers/bedtools:2.30.0--h7d7f7ad_2' }"
+
+    input:
+    tuple val(meta), path(bed)
+    val(number_of_files)
+
+    output:
+    tuple val(meta), path("*.bed"), emit: beds
+    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}"
+
+    """
+    bedtools \\
+        split \\
+        $args \\
+        -i $bed \\
+        -p $prefix \\
+        -n $number_of_files
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        bedtools: \$(bedtools --version | sed -e "s/bedtools v//g")
+    END_VERSIONS
+    """
+}
diff --git a/modules/bedtools/split/meta.yml b/modules/bedtools/split/meta.yml
new file mode 100644
index 00000000..1f41cc70
--- /dev/null
+++ b/modules/bedtools/split/meta.yml
@@ -0,0 +1,41 @@
+name: "bedtools_split"
+description: Split BED files into several smaller BED files
+keywords:
+  - sort
+tools:
+  - "bedtools":
+      description: "A powerful toolset for genome arithmetic"
+      documentation: "https://bedtools.readthedocs.io/en/latest/content/tools/sort.html"
+      licence: "['MIT', 'GPL v2']"
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - bed:
+      type: file
+      description: BED file
+      pattern: "*.bed"
+  - bed:
+      type: value
+      description: The number of files to split the BED into
+
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+  - beds:
+      type: file
+      description: list of split BED files
+      pattern: "*.bed"
+
+authors:
+  - "@nvnieuwk"
diff --git a/tests/config/pytest_modules.yml b/tests/config/pytest_modules.yml
index 142e9d41..172c3a8b 100644
--- a/tests/config/pytest_modules.yml
+++ b/tests/config/pytest_modules.yml
@@ -222,6 +222,10 @@ bedtools/sort:
   - modules/bedtools/sort/**
   - tests/modules/bedtools/sort/**
 
+bedtools/split:
+  - modules/bedtools/split/**
+  - tests/modules/bedtools/split/**
+
 bedtools/subtract:
   - modules/bedtools/subtract/**
   - tests/modules/bedtools/subtract/**
diff --git a/tests/modules/bedtools/split/main.nf b/tests/modules/bedtools/split/main.nf
new file mode 100644
index 00000000..7a1dc685
--- /dev/null
+++ b/tests/modules/bedtools/split/main.nf
@@ -0,0 +1,17 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { BEDTOOLS_SPLIT } from '../../../../modules/bedtools/split/main.nf'
+
+workflow test_bedtools_split {
+    
+    input = [
+        [ id:'test' ], // meta map
+        file(params.test_data['homo_sapiens']['genome']['genome_multi_interval_bed'], checkIfExists: true)
+    ]
+
+    number_of_files = 2
+
+    BEDTOOLS_SPLIT ( input, number_of_files )
+}
diff --git a/tests/modules/bedtools/split/nextflow.config b/tests/modules/bedtools/split/nextflow.config
new file mode 100644
index 00000000..50f50a7a
--- /dev/null
+++ b/tests/modules/bedtools/split/nextflow.config
@@ -0,0 +1,5 @@
+process {
+
+    publishDir = { "${params.outdir}/${task.process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()}" }
+    
+}
\ No newline at end of file
diff --git a/tests/modules/bedtools/split/test.yml b/tests/modules/bedtools/split/test.yml
new file mode 100644
index 00000000..92ac17fa
--- /dev/null
+++ b/tests/modules/bedtools/split/test.yml
@@ -0,0 +1,10 @@
+- name: bedtools split test_bedtools_split
+  command: nextflow run ./tests/modules/bedtools/split -entry test_bedtools_split -c ./tests/config/nextflow.config  -c ./tests/modules/bedtools/split/nextflow.config
+  tags:
+    - bedtools
+    - bedtools/split
+  files:
+    - path: output/bedtools/test.00001.bed
+      md5sum: d58e5e46c2fcc3b8be5db0f023e93cb5
+    - path: output/bedtools/test.00002.bed
+      md5sum: 03caf952e9297a54620d2bbba8dc2823