From b018e055c7fcb33ece147d92dd390a8b608a0a5d Mon Sep 17 00:00:00 2001
From: Noah Dove <nadove@ucsc.edu>
Date: Tue, 22 Nov 2022 17:44:35 -0800
Subject: [PATCH] Index `variantcallingactivity` entities (#4617)

---
 .../metadata/anvil/indexer/transform.py       |  3 ++-
 .../plugins/repository/tdr_anvil/__init__.py  | 21 +++++++++++++++++++
 2 files changed, 23 insertions(+), 1 deletion(-)

diff --git a/src/azul/plugins/metadata/anvil/indexer/transform.py b/src/azul/plugins/metadata/anvil/indexer/transform.py
index 3b56bcba35..f143315601 100644
--- a/src/azul/plugins/metadata/anvil/indexer/transform.py
+++ b/src/azul/plugins/metadata/anvil/indexer/transform.py
@@ -360,7 +360,8 @@ def _only_dataset(self) -> MutableJSON:
         'activity',
         'alignmentactivity',
         'assayactivity',
-        'sequencingactivity'
+        'sequencingactivity',
+        'variantcallingactivity'
     }
 
 
diff --git a/src/azul/plugins/repository/tdr_anvil/__init__.py b/src/azul/plugins/repository/tdr_anvil/__init__.py
index 6df72ccd89..76aa5fcc60 100644
--- a/src/azul/plugins/repository/tdr_anvil/__init__.py
+++ b/src/azul/plugins/repository/tdr_anvil/__init__.py
@@ -329,6 +329,15 @@ def _upstream_from_files(self,
                   FROM file AS f
                   JOIN {backtick(self._full_table_name(source, 'sequencingactivity'))} AS sqa
                     ON f.file_id IN UNNEST(sqa.generated_file_id)
+                UNION ALL SELECT
+                    f.file_id,
+                    'variantcallingactivity',
+                    vca.variantcallingactivity_id,
+                    vca.used_file_id,
+                    []
+                  FROM file AS f
+                  JOIN {backtick(self._full_table_name(source, 'variantcallingactivity'))} AS vca
+                    ON f.file_id IN UNNEST(vca.generated_file_id)
                 UNION ALL SELECT
                     f.file_id,
                     'activity',
@@ -419,6 +428,12 @@ def _downstream_from_files(self,
                         ala.used_file_id,
                         ala.generated_file_id
                     FROM {backtick(self._full_table_name(source, 'alignmentactivity'))} AS ala
+                    UNION ALL SELECT
+                        vca.variantcallingactivity_id,
+                        'variantcallingactivity',
+                        vca.used_file_id,
+                        vca.generated_file_id
+                    FROM {backtick(self._full_table_name(source, 'variantcallingactivity'))} AS vca
                     UNION ALL SELECT
                         a.activity_id,
                         'activity',
@@ -555,6 +570,12 @@ def convert_column(value):
             'activity_type',
             'assay_type',
             'data_modality',
+        },
+        'variantcallingactivity': {
+            'variatncallingactivity_id',
+            'activity_type',
+            'reference_assembly',
+            'data_modality'
         }
     }