From a14855dbe8648ca445759f57a1a2805ec4812615 Mon Sep 17 00:00:00 2001
From: James Fellows Yates <jfy133@gmail.com>
Date: Thu, 2 Jun 2022 21:22:21 +0200
Subject: [PATCH] Start meta

---
 modules/gatk/unifiedgenotyper/main.nf  | 16 ++++++-
 modules/gatk/unifiedgenotyper/meta.yml | 59 ++++++++++++++++----------
 2 files changed, 51 insertions(+), 24 deletions(-)

diff --git a/modules/gatk/unifiedgenotyper/main.nf b/modules/gatk/unifiedgenotyper/main.nf
index 4dd6e408..896a347e 100644
--- a/modules/gatk/unifiedgenotyper/main.nf
+++ b/modules/gatk/unifiedgenotyper/main.nf
@@ -13,9 +13,13 @@ process GATK_UNIFIEDGENOTYPER {
     path(fai)
     path(dict)
     path(known_vcf)
+    path(intervals)
+    path(contamination)
+    path(dbsnps)
+    path(comp)
 
     output:
-    tuple val(meta), path("*.bam"), emit: bam
+    tuple val(meta), path("*.vcf.gz"), emit: vcf
     path "versions.yml"           , emit: versions
 
     when:
@@ -24,6 +28,10 @@ process GATK_UNIFIEDGENOTYPER {
     script:
     def args = task.ext.args ?: ''
     def prefix = task.ext.prefix ?: "${meta.id}"
+    def contamination_file = contamination ? "-contaminationFile ${contamination}" : ""
+    def dbsnps_file = dbsnps ? "--dbsnp ${dbsnps}" : ""
+    def comp_file = comp ? "--comp ${comp}" : ""
+    def intervals_file = intervals ? "--intervals ${intervals}" : ""
 
     def avail_mem = 3
     if (!task.memory) {
@@ -39,9 +47,15 @@ process GATK_UNIFIEDGENOTYPER {
         -T UnifiedGenotyper \\
         -I ${input} \\
         -R ${fasta} \\
+        ${contamination_file} \\
+        ${dbsnps_file} \\
+        ${comp_file} \\
+        ${intervals_file}
         -o ${prefix}.vcf \\
         $args
 
+    gzip -n *.vcf
+
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":
         gatk: \$(echo \$(samtools --version 2>&1) | sed 's/^.*samtools //; s/Using.*\$//' ))
diff --git a/modules/gatk/unifiedgenotyper/meta.yml b/modules/gatk/unifiedgenotyper/meta.yml
index 7a992591..5f85c451 100644
--- a/modules/gatk/unifiedgenotyper/meta.yml
+++ b/modules/gatk/unifiedgenotyper/meta.yml
@@ -1,51 +1,64 @@
 name: "gatk_unifiedgenotyper"
-## TODO nf-core: Add a description of the module and list keywords
-description: write your description here
 keywords:
-  - sort
+  - bam
+  - vcf
+  - variant calling
+  - indel
+  - realignment
+  - targets
 tools:
   - "gatk":
-      ## TODO nf-core: Add a description and other details for the software below
       description: "The full Genome Analysis Toolkit (GATK) framework, license restricted."
-      homepage: "None"
-      documentation: "None"
-      tool_dev_url: "None"
-      doi: ""
-      licence: "['BSD', 'https://www.broadinstitute.org/gatk/about/#licensing', 'https://software.broadinstitute.org/gatk/download/licensing']"
+      homepage: "https://gatk.broadinstitute.org/hc/en-us"
+      documentation: "https://github.com/broadinstitute/gatk-docs"
+      licence: "['https://software.broadinstitute.org/gatk/download/licensing', 'BSD', 'https://www.broadinstitute.org/gatk/about/#licensing']"
 
-## TODO nf-core: Add a description of all of the variables used as input
 input:
-  # Only when we have meta
   - meta:
       type: map
       description: |
         Groovy Map containing sample information
         e.g. [ id:'test', single_end:false ]
-  # 
-  ## TODO nf-core: Delete / customise this example input
-  - bam:
+  - input:
       type: file
-      description: BAM/CRAM/SAM file
-      pattern: "*.{bam,cram,sam}"
+      description: Sorted and indexed BAM/CRAM/SAM file
+      pattern: "*.bam"
+  - index:
+      type: file
+      description: BAM index file
+      pattern: "*.bai"
+  - fasta:
+      type: file
+      description: Reference file used to generate BAM file
+      pattern: ".{fasta,fa,fna}"
+  - fai:
+      type: file
+      description: Index of reference file used to generate BAM file
+      pattern: ".fai"
+  - dict:
+      type: file
+      description: GATK dict file for reference
+      pattern: ".dict"
+  - known_vcf:
+      type: file
+      description: Optional input VCF file(s) with known indels
+      pattern: ".vcf"
 
-## TODO nf-core: Add a description of all of the variables used as output
 output:
-  #Only when we have meta
   - 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"
-  ## TODO nf-core: Delete / customise this example output
-  - bam:
+  - vcf:
       type: file
-      description: Sorted BAM/CRAM/SAM file
-      pattern: "*.{bam,cram,sam}"
+      description: VCF file containing called variants
+      pattern: "*.vcf.gz"
 
 authors:
+  - "@ilight1542"
   - "@jfy133"