VS-1549 Add VAT to integration tests

.dockstore.yml

@@ -316,6 +316,17 @@ workflows:
          branches:
              - master
              - ah_var_store
+             - gg_VS-1549_AddVATToIntegrationTests
+         tags:
+             - /.*/
+   - name: GvsQuickstartVATIntegration
+     subclass: WDL
+     primaryDescriptorPath: /scripts/variantstore/wdl/test/GvsQuickstartVATIntegration.wdl
+     filters:
+         branches:
+             - master
+             - ah_var_store
+             - gg_VS-1549_AddVATToIntegrationTests
          tags:
              - /.*/
    - name: GvsIngestTieout

scripts/variantstore/variant-annotations-table/GvsCreateVATfromVDS.wdl

@@ -85,7 +85,7 @@ workflow GvsCreateVATfromVDS {
 
     # If the vat version is undefined or v1 then the vat tables would be named like filter_vat, otherwise filter_vat_v2.
     String effective_vat_version = if (defined(vat_version) && select_first([vat_version]) != "v1") then "_" + select_first([vat_version]) else ""
-    String vat_table_name = filter_set_name + "_vat" + effective_vat_version
+    String effective_vat_table_name = filter_set_name + "_vat" + effective_vat_version
 
     String output_path_without_a_trailing_slash = sub(output_path, "/$", "")
     String effective_output_path = if (output_path == output_path_without_a_trailing_slash) then output_path + "/" else output_path
@@ -262,7 +262,7 @@ workflow GvsCreateVATfromVDS {
                 project_id = project_id,
                 dataset_name = dataset_name,
                 output_path = effective_output_path,
-                base_vat_table_name = vat_table_name,
+                base_vat_table_name = effective_vat_table_name,
                 prep_vt_json_done = PrepVtAnnotationJson.done,
                 prep_genes_json_done = PrepGenesAnnotationJson.done,
                 cloud_sdk_docker = effective_cloud_sdk_docker,
@@ -271,7 +271,7 @@ workflow GvsCreateVATfromVDS {
         call DeduplicateVatInBigQuery {
             input:
                 input_vat_table_name = BigQueryLoadJson.vat_table,
-                output_vat_table_name = vat_table_name,
+                output_vat_table_name = effective_vat_table_name,
                 nirvana_schema = MakeSubpopulationFilesAndReadSchemaFiles.vat_schema_json_file,
                 project_id = project_id,
                 dataset_name = dataset_name,
@@ -294,6 +294,7 @@ workflow GvsCreateVATfromVDS {
     }
 
     output {
+        String vat_table_name = effective_vat_table_name
         String? cluster_name = GenerateSitesOnlyVcf.cluster_name
         File? dropped_sites_file = MergeTsvs.output_file
         File? final_tsv_file = GvsCreateVATFilesFromBigQuery.final_tsv_file

scripts/variantstore/variant-annotations-table/GvsValidateVAT.wdl

@@ -8,6 +8,7 @@ workflow GvsValidateVat {
         String project_id
         String dataset_name
         String vat_table_name
+        Boolean? is_small_callset
         String? cloud_sdk_docker
         String? variants_docker
     }
@@ -25,20 +26,23 @@ workflow GvsValidateVat {
     String effective_cloud_sdk_docker = select_first([cloud_sdk_docker, GetToolVersions.cloud_sdk_docker])
     String effective_variants_docker = select_first([variants_docker, GetToolVersions.variants_docker])
 
-    call Utils.GetBQTableLastModifiedDatetime as SampleDateTime {
-        input:
-        project_id = project_id,
-        fq_table = fq_vat_table,
-        cloud_sdk_docker = effective_cloud_sdk_docker,
-    }
+    # Definining is_small_callset allows us to run this WDL on a dataset that has not had samples loaded (for testing)
+    if (!defined(is_small_callset)) {
+        call Utils.GetBQTableLastModifiedDatetime as SampleDateTime {
+            input:
+                project_id = project_id,
+                fq_table = fq_sample_table,
+                cloud_sdk_docker = effective_cloud_sdk_docker,
+        }
 
-    call Utils.GetNumSamplesLoaded {
-        input:
-            fq_sample_table = fq_sample_table,
-            project_id = project_id,
-            sample_table_timestamp = SampleDateTime.last_modified_timestamp,
-            control_samples = false,
-            cloud_sdk_docker = effective_cloud_sdk_docker,
+        call Utils.GetNumSamplesLoaded {
+            input:
+                fq_sample_table = fq_sample_table,
+                project_id = project_id,
+                sample_table_timestamp = SampleDateTime.last_modified_timestamp,
+                control_samples = false,
+                cloud_sdk_docker = effective_cloud_sdk_docker,
+        }
     }
 
     call Utils.GetBQTableLastModifiedDatetime as VatDateTime {
@@ -153,7 +157,7 @@ workflow GvsValidateVat {
     }
 
     # only check certain things if the callset is larger than 10,000 samples (a guess)
-    Boolean callset_is_small = GetNumSamplesLoaded.num_samples < 10000
+    Boolean callset_is_small = select_first([is_small_callset, select_first([GetNumSamplesLoaded.num_samples, 1]) < 10000])
     if (!callset_is_small) {
         call ClinvarSignificance {
             input:

scripts/variantstore/wdl/test/GvsQuickstartIntegration.wdl

@@ -2,6 +2,7 @@ version 1.0
 
 import "GvsQuickstartVcfIntegration.wdl" as QuickstartVcfIntegration
 import "GvsQuickstartHailIntegration.wdl" as QuickstartHailIntegration
+import "GvsQuickstartVATIntegration.wdl" as QuickstartVATIntegration
 import "../GvsJointVariantCalling.wdl" as JointVariantCalling
 import "../GvsUtils.wdl" as Utils
 
@@ -14,6 +15,8 @@ workflow GvsQuickstartIntegration {
         Boolean run_exome_integration = true
         Boolean run_beta_integration = true
         Boolean run_bge_integration = true
+        Boolean run_vat_integration = true
+        Boolean run_vat_integration_test_from_vds = true        # If false, will use sites-only VCF
         String sample_id_column_name = "sample_id"
         String vcf_files_column_name = "hg38_reblocked_gvcf"
         String vcf_index_files_column_name = "hg38_reblocked_gvcf_index"
@@ -25,6 +28,7 @@ workflow GvsQuickstartIntegration {
         String? cloud_sdk_docker
         String? cloud_sdk_slim_docker
         String? variants_docker
+        String? variants_nirvana_docker
         String? gatk_docker
         String? hail_version
         Boolean chr20_X_Y_only = true
@@ -52,6 +56,7 @@ workflow GvsQuickstartIntegration {
     String effective_cloud_sdk_docker = select_first([cloud_sdk_docker, GetToolVersions.cloud_sdk_docker])
     String effective_cloud_sdk_slim_docker = select_first([cloud_sdk_slim_docker, GetToolVersions.cloud_sdk_slim_docker])
     String effective_variants_docker = select_first([variants_docker, GetToolVersions.variants_docker])
+    String effective_variants_nirvana_docker = select_first([variants_nirvana_docker, GetToolVersions.variants_nirvana_docker])
     String effective_gatk_docker = select_first([gatk_docker, GetToolVersions.gatk_docker])
     String effective_hail_version = select_first([hail_version, GetToolVersions.hail_version])
 
@@ -72,6 +77,10 @@ workflow GvsQuickstartIntegration {
         }
     }
 
+    String workspace_bucket = GetToolVersions.workspace_bucket
+    String workspace_id = GetToolVersions.workspace_id
+    String submission_id = GetToolVersions.submission_id
+
     # Note for `GvsQuickstartIntegration` we use the git_branch_or_tag *input* and its corresponding git hash. This is not
     # necessarily the same as the branch name selected in Terra for the integration `GvsQuickstartIntegration` workflow,
     # though in practice likely they are the same.
@@ -98,9 +107,9 @@ workflow GvsQuickstartIntegration {
                 cloud_sdk_slim_docker = effective_cloud_sdk_slim_docker,
                 variants_docker = effective_variants_docker,
                 gatk_docker = effective_gatk_docker,
-                workspace_bucket = GetToolVersions.workspace_bucket,
-                workspace_id = GetToolVersions.workspace_id,
-                submission_id = GetToolVersions.submission_id,
+                workspace_bucket = workspace_bucket,
+                workspace_id = workspace_id,
+                submission_id = submission_id,
                 hail_version = effective_hail_version,
                 maximum_alternate_alleles = maximum_alternate_alleles,
         }
@@ -137,9 +146,9 @@ workflow GvsQuickstartIntegration {
                 cloud_sdk_slim_docker = effective_cloud_sdk_slim_docker,
                 variants_docker = effective_variants_docker,
                 gatk_docker = effective_gatk_docker,
-                workspace_bucket = GetToolVersions.workspace_bucket,
-                workspace_id = GetToolVersions.workspace_id,
-                submission_id = GetToolVersions.submission_id,
+                workspace_bucket = workspace_bucket,
+                workspace_id = workspace_id,
+                submission_id = submission_id,
                 maximum_alternate_alleles = maximum_alternate_alleles,
         }
         call QuickstartVcfIntegration.GvsQuickstartVcfIntegration as QuickstartVcfVQSRIntegration {
@@ -164,9 +173,9 @@ workflow GvsQuickstartIntegration {
                 cloud_sdk_slim_docker = effective_cloud_sdk_slim_docker,
                 variants_docker = effective_variants_docker,
                 gatk_docker = effective_gatk_docker,
-                workspace_bucket = GetToolVersions.workspace_bucket,
-                workspace_id = GetToolVersions.workspace_id,
-                submission_id = GetToolVersions.submission_id,
+                workspace_bucket = workspace_bucket,
+                workspace_id = workspace_id,
+                submission_id = submission_id,
                 maximum_alternate_alleles = maximum_alternate_alleles,
         }
 
@@ -210,9 +219,9 @@ workflow GvsQuickstartIntegration {
                 cloud_sdk_slim_docker = effective_cloud_sdk_slim_docker,
                 variants_docker = effective_variants_docker,
                 gatk_docker = effective_gatk_docker,
-                workspace_bucket = GetToolVersions.workspace_bucket,
-                workspace_id = GetToolVersions.workspace_id,
-                submission_id = GetToolVersions.submission_id,
+                workspace_bucket = workspace_bucket,
+                workspace_id = workspace_id,
+                submission_id = submission_id,
                 maximum_alternate_alleles = maximum_alternate_alleles,
                 target_interval_list = target_interval_list,
         }
@@ -249,9 +258,9 @@ workflow GvsQuickstartIntegration {
                 cloud_sdk_slim_docker = effective_cloud_sdk_slim_docker,
                 variants_docker = effective_variants_docker,
                 gatk_docker = effective_gatk_docker,
-                workspace_bucket = GetToolVersions.workspace_bucket,
-                workspace_id = GetToolVersions.workspace_id,
-                submission_id = GetToolVersions.submission_id,
+                workspace_bucket = workspace_bucket,
+                workspace_id = workspace_id,
+                submission_id = submission_id,
                 maximum_alternate_alleles = maximum_alternate_alleles,
                 target_interval_list = target_interval_list,
         }
@@ -268,8 +277,6 @@ workflow GvsQuickstartIntegration {
     if (run_beta_integration) {
         String project_id = "gvs-internal"
 
-        String workspace_bucket = GetToolVersions.workspace_bucket
-        String submission_id = GetToolVersions.submission_id
         String extract_output_gcs_dir = "~{workspace_bucket}/output_vcfs/by_submission_id/~{submission_id}/beta"
         Boolean collect_variant_calling_metrics = true
 
@@ -296,9 +303,9 @@ workflow GvsQuickstartIntegration {
                 cloud_sdk_docker = effective_cloud_sdk_docker,
                 variants_docker = effective_variants_docker,
                 gatk_docker = effective_gatk_docker,
-                workspace_bucket = GetToolVersions.workspace_bucket,
-                workspace_id = GetToolVersions.workspace_id,
-                submission_id = GetToolVersions.submission_id,
+                workspace_bucket = workspace_bucket,
+                workspace_id = workspace_id,
+                submission_id = submission_id,
                 maximum_alternate_alleles = maximum_alternate_alleles,
                 git_branch_or_tag = git_branch_or_tag,
                 sample_id_column_name = sample_id_column_name,
@@ -317,6 +324,27 @@ workflow GvsQuickstartIntegration {
         }
     }
 
+    if (run_vat_integration) {
+        String extract_vat_output_gcs_dir = "~{workspace_bucket}/output_vat/by_submission_id/~{submission_id}/vat"
+
+        call QuickstartVATIntegration.GvsQuickstartVATIntegration as GvsQuickstartVATIntegration {
+            input:
+                git_branch_or_tag = git_branch_or_tag,
+                git_hash = GetToolVersions.git_hash,
+                use_default_dockers = use_default_dockers,
+                expected_output_prefix = expected_output_prefix,
+                dataset_suffix = "vat",
+                output_path = extract_vat_output_gcs_dir,
+                use_vds = run_vat_integration_test_from_vds,
+                basic_docker = effective_basic_docker,
+                cloud_sdk_docker = effective_cloud_sdk_docker,
+                cloud_sdk_slim_docker = effective_cloud_sdk_slim_docker,
+                variants_docker = effective_variants_docker,
+                variants_nirvana_docker = effective_variants_nirvana_docker,
+                gatk_docker = effective_gatk_docker,
+        }
+    }
+
     output {
         String recorded_git_hash = GetToolVersions.git_hash
     }