From 8c8be7d7c6c5b17b89f1c392866fc433b0f79d69 Mon Sep 17 00:00:00 2001
From: FriederikeHanssen <Friederike.hanssen@qbic.uni-tuebingen.de>
Date: Thu, 17 Feb 2022 15:05:21 +0100
Subject: [PATCH] Add gatherpileupsummaries (#1311)

* Add gatherpileupsummaries

* fix checksum

* Update modules/gatk4/gatherpileupsummaries/main.nf

* Add in when
---
 modules/gatk4/gatherpileupsummaries/main.nf   | 45 +++++++++++++++++++
 modules/gatk4/gatherpileupsummaries/meta.yml  | 41 +++++++++++++++++
 tests/config/pytest_modules.yml               |  4 ++
 .../gatk4/gatherpileupsummaries/main.nf       | 18 ++++++++
 .../gatherpileupsummaries/nextflow.config     |  5 +++
 .../gatk4/gatherpileupsummaries/test.yml      |  8 ++++
 6 files changed, 121 insertions(+)
 create mode 100644 modules/gatk4/gatherpileupsummaries/main.nf
 create mode 100644 modules/gatk4/gatherpileupsummaries/meta.yml
 create mode 100644 tests/modules/gatk4/gatherpileupsummaries/main.nf
 create mode 100644 tests/modules/gatk4/gatherpileupsummaries/nextflow.config
 create mode 100644 tests/modules/gatk4/gatherpileupsummaries/test.yml

diff --git a/modules/gatk4/gatherpileupsummaries/main.nf b/modules/gatk4/gatherpileupsummaries/main.nf
new file mode 100644
index 00000000..52e57127
--- /dev/null
+++ b/modules/gatk4/gatherpileupsummaries/main.nf
@@ -0,0 +1,45 @@
+process GATK4_GATHERPILEUPSUMMARIES {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda (params.enable_conda ? "bioconda::gatk4=4.2.5.0" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/gatk4:4.2.5.0--hdfd78af_0' :
+        'quay.io/biocontainers/gatk4:4.2.5.0--hdfd78af_0' }"
+
+
+    input:
+    tuple val(meta), path(pileup)
+    path dict
+
+    output:
+    tuple val(meta), path("*.pileupsummaries.table"), emit: table
+    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 input = pileup.collect{ "-I ${it} " }.join(' ')
+
+    def avail_mem = 3
+    if (!task.memory) {
+        log.info '[GATK GatherPileupSummaries] Available memory not known - defaulting to 3GB. Specify process memory requirements to change this.'
+    } else {
+        avail_mem = task.memory.giga
+    }
+    """
+    gatk --java-options "-Xmx${avail_mem}g" \
+        GatherPileupSummaries \
+        --sequence-dictionary ${dict} \
+        ${input} \
+        -O ${prefix}.pileupsummaries.table
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        gatk4: \$(echo \$(gatk --version 2>&1) | sed 's/^.*(GATK) v//; s/ .*\$//')
+    END_VERSIONS
+    """
+}
diff --git a/modules/gatk4/gatherpileupsummaries/meta.yml b/modules/gatk4/gatherpileupsummaries/meta.yml
new file mode 100644
index 00000000..7885a930
--- /dev/null
+++ b/modules/gatk4/gatherpileupsummaries/meta.yml
@@ -0,0 +1,41 @@
+name: gatk4_gatherpileupsummaries
+description: write your description here
+keywords:
+  - sort
+tools:
+  - gatk4:
+      description: Genome Analysis Toolkit (GATK4)
+      homepage: https://gatk.broadinstitute.org/hc/en-us
+      documentation: https://gatk.broadinstitute.org/hc/en-us
+      tool_dev_url: https://github.com/broadinstitute/gatk
+      doi: "10.1158/1538-7445.AM2017-3590"
+      licence: ['BSD-3-clause']
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - pileup:
+      type: (list of) file(s)
+      description: Pileup files from gatk4/getpileupsummaries
+      pattern: "*.pileups.table"
+
+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"
+  - table:
+      type: file
+      description: Pileup file
+      pattern: "*.pileups.table"
+
+authors:
+  - "@FriederikeHanssen"
diff --git a/tests/config/pytest_modules.yml b/tests/config/pytest_modules.yml
index 1b855da9..3825a873 100644
--- a/tests/config/pytest_modules.yml
+++ b/tests/config/pytest_modules.yml
@@ -580,6 +580,10 @@ gatk4/gatherbqsrreports:
   - modules/gatk4/gatherbqsrreports/**
   - tests/modules/gatk4/gatherbqsrreports/**
 
+gatk4/gatherpileupsummaries:
+  - modules/gatk4/gatherpileupsummaries/**
+  - tests/modules/gatk4/gatherpileupsummaries/**
+
 gatk4/genomicsdbimport:
   - modules/gatk4/genomicsdbimport/**
   - tests/modules/gatk4/genomicsdbimport/**
diff --git a/tests/modules/gatk4/gatherpileupsummaries/main.nf b/tests/modules/gatk4/gatherpileupsummaries/main.nf
new file mode 100644
index 00000000..2bd7e98c
--- /dev/null
+++ b/tests/modules/gatk4/gatherpileupsummaries/main.nf
@@ -0,0 +1,18 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { GATK4_GATHERPILEUPSUMMARIES } from '../../../../modules/gatk4/gatherpileupsummaries/main.nf'
+
+workflow test_gatk4_gatherpileupsummaries {
+
+    input = [
+        [ id:'test', single_end:false ], // meta map
+        [file(params.test_data['homo_sapiens']['illumina']['test_pileups_table'], checkIfExists: true)]
+        //file(params.test_data['homo_sapiens']['illumina']['test_pileups_table'], checkIfExists: true)]
+    ]
+
+    dict = file(params.test_data['homo_sapiens']['genome']['genome_21_dict'], checkIfExists: true)
+
+    GATK4_GATHERPILEUPSUMMARIES ( input, dict )
+}
diff --git a/tests/modules/gatk4/gatherpileupsummaries/nextflow.config b/tests/modules/gatk4/gatherpileupsummaries/nextflow.config
new file mode 100644
index 00000000..50f50a7a
--- /dev/null
+++ b/tests/modules/gatk4/gatherpileupsummaries/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/gatk4/gatherpileupsummaries/test.yml b/tests/modules/gatk4/gatherpileupsummaries/test.yml
new file mode 100644
index 00000000..0c38a602
--- /dev/null
+++ b/tests/modules/gatk4/gatherpileupsummaries/test.yml
@@ -0,0 +1,8 @@
+- name: gatk4 gatherpileupsummaries
+  command: nextflow run ./tests/modules/gatk4/gatherpileupsummaries -entry test_gatk4_gatherpileupsummaries -c ./tests/config/nextflow.config -c ./tests/modules/gatk4/gatherpileupsummaries/nextflow.config
+  tags:
+    - gatk4
+    - gatk4/gatherpileupsummaries
+  files:
+    - path: output/gatk4/test.pileupsummaries.table
+      md5sum: 8e0ca6f66e112bd2f7ec1d31a2d62469