Add VAT clinvar reverse complement

.dockstore.yml

@@ -211,7 +211,7 @@ workflows:
          branches:
              - master
              - ah_var_store
-             - rc-1528-n-rounds-update
+             - rc-vs-1457-vat-clinvar
          tags:
              - /.*/
    - name: GvsCreateVATFilesFromBigQuery

scripts/variantstore/scripts/create_vt_bqloadjson_from_annotations.py

@@ -6,6 +6,7 @@
 import argparse
 import logging
 import sys
+import Bio.Seq
 
 vat_nirvana_positions_dictionary = {
     "position": "position", # required
@@ -233,14 +234,19 @@ def make_annotated_json_row(row_position, row_ref, row_alt, variant_line, transc
         updated_dates = [] # grab the most recent
         phenotypes = [] # ordered alphabetically
         clinvar_ids = [] # For easy validation downstream
-        # Note that inside the clinvar array, are multiple objects that may or may not be the one we are looking for. We check by making sure the ref and alt are the same
+        # Note that inside the clinvar array, are multiple objects that may or may not be the one we are looking for.
+        # We check by making sure the ref and alt are the same including any reverse complements
+        ## TODO add clinvar star rating!!!
         for clinvar_obj in clinvar_objs:
             # get only the clinvar objs with right variant and the id that starts with RCV
-            if (clinvar_obj.get("refAllele") == var_ref) and (clinvar_obj.get("altAllele") == var_alt) and (clinvar_obj.get("id")[:3] == "RCV"):
+            if (((clinvar_obj.get("refAllele") == var_ref) and (clinvar_obj.get("altAllele") == var_alt)) or
+            ((clinvar_obj.get("refAllele") == Bio.Seq.reverse_complement(variant_line["refAllele"])) and (clinvar_obj.get("altAllele") == Bio.Seq.reverse_complement(variant_line["altAllele"])))) and (clinvar_obj.get("id")[:3] == "RCV"):
                 clinvar_ids.append(clinvar_obj.get("id"))
                 significance_values.extend([x.lower() for x in clinvar_obj.get("significance")])
                 updated_dates.append(clinvar_obj.get("lastUpdatedDate"))
                 phenotypes.extend(clinvar_obj.get("phenotypes"))
+                ## TODO add the ("variationId") and the ("reviewStatus")--note that the reviewStatus will need to maintain the ordering of the significance arrays
+                ## we need to do this with a tuple so that the reviewStatus lines up with the significance (since significance seems to be an array, while star is a single value)
         if len(clinvar_ids) > 0:
             ordered_significance_values = []
             # We want to collect all the significance values and order them by the significance_ordering list

scripts/variantstore/scripts/requirements.txt

@@ -5,3 +5,4 @@ google-cloud-storage
 firecloud
 terra-notebook-utils
 pybedtools
+biopython

scripts/variantstore/wdl/GvsUtils.wdl

@@ -72,7 +72,7 @@ task GetToolVersions {
     # GVS generally uses the smallest `alpine` version of the Google Cloud SDK as it suffices for most tasks, but
     # there are a handlful of tasks that require the larger GNU libc-based `slim`.
     String cloud_sdk_slim_docker = "gcr.io/google.com/cloudsdktool/cloud-sdk:435.0.0-slim"
-    String variants_docker = "us-central1-docker.pkg.dev/broad-dsde-methods/gvs/variants:2024-11-25-alpine-913039adf8f4"
+    String variants_docker = "us-central1-docker.pkg.dev/broad-dsde-methods/gvs/variants:2025-01-17-alpine-2b4c2ca0187c"
     String variants_nirvana_docker = "us.gcr.io/broad-dsde-methods/variantstore:nirvana_2022_10_19"
     String gatk_docker = "us-central1-docker.pkg.dev/broad-dsde-methods/gvs/gatk:2024-11-24-gatkbase-1807487d5912"
     String real_time_genomics_docker = "docker.io/realtimegenomics/rtg-tools:latest"