PGEN pvar files not compressed [VS-1412] (#8872)

broadinstitute · Jun 17, 2024 · 862e4a5 · 862e4a5
1 parent 9dc13be
commit 862e4a5
Show file tree

Hide file tree

Showing 2 changed files with 12 additions and 3 deletions.
diff --git a/.dockstore.yml b/.dockstore.yml
@@ -459,6 +459,7 @@ workflows:
          - master
          - ah_var_store
          - EchoCallset
+         - vs_1412_pgen_pvars_not_compressed
    - name: MergePgenHierarchicalWdl
      subclass: WDL
      primaryDescriptorPath: /scripts/variantstore/wdl/MergePgenHierarchical.wdl

diff --git a/scripts/variantstore/wdl/MergePgenHierarchical.wdl b/scripts/variantstore/wdl/MergePgenHierarchical.wdl
@@ -85,7 +85,9 @@ task MergePgen {
     Int cpu = threads + 1
 
     command <<<
-        set -euxo pipefail
+        # Prepend date, time and pwd to xtrace log entries.
+        PS4='\D{+%F %T} \w $ '
+        set -o errexit -o nounset -o pipefail -o xtrace
 
         # Download files using gsutil
         mkdir pgen_dir
@@ -109,7 +111,7 @@ task MergePgen {
         0)
             echo "No pgen files so creating empty ones"
             touch ~{output_file_base_name}.pgen
-            touch ~{output_file_base_name}.pvar
+            touch ~{output_file_base_name}.pvar.zst
             touch ~{output_file_base_name}.psam
             ;;
         1)
@@ -135,11 +137,17 @@ task MergePgen {
             ;;
         esac
 
+        if [[ -e ~{output_file_base_name}.pvar.zst ]]
+        then
+            # Decompress for earlier versions of PLINK that can't handle zst compression.
+            plink2 --zst-decompress ~{output_file_base_name}.pvar.zst > ~{output_file_base_name}.pvar
+        fi
+
     >>>
 
     output {
         File pgen_file = "${output_file_base_name}.pgen"
-        File pvar_file = "${output_file_base_name}.pvar.zst"
+        File pvar_file = "${output_file_base_name}.pvar"
         File psam_file = "${output_file_base_name}.psam"
     }