From 8e5eaf1fa442225f8d24e95bba2a19a7e2fc9ce9 Mon Sep 17 00:00:00 2001
From: Benjamin Wingfield <bwingfield@ebi.ac.uk>
Date: Wed, 2 Mar 2022 00:43:28 +0000
Subject: [PATCH] Update plink2/vcf to output zstandard compressed data
 automatically (#1258)

* update plink2/vcf to output zstandard compressed data automatically

* update meta

* set plink CPU and memory usage

Co-authored-by: Sateesh <33637490+sateeshperi@users.noreply.github.com>
---
 modules/plink2/vcf/main.nf        | 12 ++++++++----
 modules/plink2/vcf/meta.yml       |  2 +-
 tests/modules/plink2/vcf/test.yml | 10 ++++++----
 3 files changed, 15 insertions(+), 9 deletions(-)

diff --git a/modules/plink2/vcf/main.nf b/modules/plink2/vcf/main.nf
index 041cb4b9..090851a8 100644
--- a/modules/plink2/vcf/main.nf
+++ b/modules/plink2/vcf/main.nf
@@ -11,10 +11,10 @@ process PLINK2_VCF {
     tuple val(meta), path(vcf)
 
     output:
-    tuple val(meta), path("*.pgen"), emit: pgen
-    tuple val(meta), path("*.psam"), emit: psam
-    tuple val(meta), path("*.pvar"), emit: pvar
-    path "versions.yml"            , emit: versions
+    tuple val(meta), path("*.pgen")    , emit: pgen
+    tuple val(meta), path("*.psam")    , emit: psam
+    tuple val(meta), path("*.pvar.zst"), emit: pvar
+    path "versions.yml"                , emit: versions
 
     when:
     task.ext.when == null || task.ext.when
@@ -22,10 +22,14 @@ process PLINK2_VCF {
     script:
     def args = task.ext.args ?: ''
     def prefix = task.ext.prefix ?: "${meta.id}"
+    def mem_mb = task.memory.toMega()
     """
     plink2 \\
+        --threads $task.cpus \\
+        --memory $mem_mb \\
         $args \\
         --vcf $vcf \\
+        --make-pgen vzs \\
         --out ${prefix}
 
     cat <<-END_VERSIONS > versions.yml
diff --git a/modules/plink2/vcf/meta.yml b/modules/plink2/vcf/meta.yml
index 5697ebef..3c776daf 100644
--- a/modules/plink2/vcf/meta.yml
+++ b/modules/plink2/vcf/meta.yml
@@ -46,7 +46,7 @@ output:
   - pvar:
       type: file
       description: PLINK 2 variant information file
-      pattern: "*.{psam}"
+      pattern: "*.{pvar.zst}"
 
 authors:
   - "@nebfield"
diff --git a/tests/modules/plink2/vcf/test.yml b/tests/modules/plink2/vcf/test.yml
index 52f58a42..d354af48 100644
--- a/tests/modules/plink2/vcf/test.yml
+++ b/tests/modules/plink2/vcf/test.yml
@@ -1,12 +1,14 @@
 - name: plink2 vcf test_plink2_vcf
-  command: nextflow run ./tests/modules/plink2/vcf -entry test_plink2_vcf -c ./tests/config/nextflow.config -c ./tests/modules/plink2/vcf/nextflow.config
+  command: nextflow run tests/modules/plink2/vcf -entry test_plink2_vcf -c tests/config/nextflow.config
   tags:
-    - plink2/vcf
     - plink2
+    - plink2/vcf
   files:
     - path: output/plink2/test.pgen
       md5sum: d66d3cd4a6c9cca1a4073d7f4b277041
     - path: output/plink2/test.psam
       md5sum: dc3b77d7753a7bed41734323e3549b10
-    - path: output/plink2/test.pvar
-      md5sum: d61e53f847a6335138b584216b4e45d0
+    - path: output/plink2/test.pvar.zst
+      md5sum: b53cccb83e024a39789af5eab8de1c28
+    - path: output/plink2/versions.yml
+      md5sum: 82ada74bc81473b7cba377f696acf54c