fix: enhance error handling in get_allele_registry_associations function

Author: bencap

src/mavedb/lib/clingen/services.py

@@ -1,19 +1,17 @@
 import hashlib
 import logging
-import requests
 import os
 import time
 from datetime import datetime
-from typing import Optional
+from typing import Optional, Union
 from urllib import parse
 
-
+import requests
 from jose import jwt
 
-from mavedb.lib.logging.context import logging_context, save_to_logging_context, format_raised_exception_info_as_dict
 from mavedb.lib.clingen.constants import GENBOREE_ACCOUNT_NAME, GENBOREE_ACCOUNT_PASSWORD, LDH_MAVE_ACCESS_ENDPOINT
-
-from mavedb.lib.types.clingen import LdhSubmission, ClinGenAllele
+from mavedb.lib.logging.context import format_raised_exception_info_as_dict, logging_context, save_to_logging_context
+from mavedb.lib.types.clingen import ClinGenAllele, ClinGenSubmissionError, LdhSubmission
 from mavedb.lib.utils import batched
 
 logger = logging.getLogger(__name__)
@@ -71,7 +69,9 @@ def construct_auth_url(self, url: str) -> str:
         token = hashlib.sha1((url + identity + gbTime).encode("utf-8")).hexdigest()
         return url + "&gbLogin=" + GENBOREE_ACCOUNT_NAME + "&gbTime=" + gbTime + "&gbToken=" + token
 
-    def dispatch_submissions(self, content_submissions: list[str]) -> list[ClinGenAllele]:
+    def dispatch_submissions(
+        self, content_submissions: list[str]
+    ) -> list[Union[ClinGenAllele, ClinGenSubmissionError]]:
         save_to_logging_context({"car_submission_count": len(content_submissions)})
 
         try:
@@ -89,7 +89,7 @@ def dispatch_submissions(self, content_submissions: list[str]) -> list[ClinGenAl
             logger.error(msg="Failed to dispatch CAR submission.", exc_info=exc, extra=logging_context())
             return []
 
-        response_data: list[ClinGenAllele] = response.json()
+        response_data: list[Union[ClinGenAllele, ClinGenSubmissionError]] = response.json()
         save_to_logging_context({"car_submission_response_count": len(response_data)})
         logger.info(msg="Successfully dispatched CAR submission.", extra=logging_context())
 
@@ -324,7 +324,7 @@ def clingen_allele_id_from_ldh_variation(variation: Optional[dict]) -> Optional[
 
 
 def get_allele_registry_associations(
-    content_submissions: list[str], submission_response: list[ClinGenAllele]
+    content_submissions: list[str], submission_response: list[Union[ClinGenAllele, ClinGenSubmissionError]]
 ) -> dict[str, str]:
     """
     Links HGVS strings and ClinGen Canonoical Allele IDs (CAIDs) given a list of both.
@@ -360,6 +360,13 @@ def get_allele_registry_associations(
 
     allele_registry_associations: dict[str, str] = {}
     for registration in submission_response:
+        if "errorType" in registration:
+            logger.warning(
+                msg=f"Skipping errored ClinGen Allele Registry HGVS {registration.get('hgvs', 'unknown')} ({registration.get('errorType', 'unknown')}): {registration.get('message', 'unknown error message')}",
+                extra=logging_context(),
+            )
+            continue
+
         # Extract the CAID from the URL (e.g., "http://reg.test.genome.network/allele/CA2513066" -> "CA2513066")
         caid = registration["@id"].split("/")[-1]
         alleles = registration.get("genomicAlleles", []) + registration.get("transcriptAlleles", [])

src/mavedb/lib/types/clingen.py

@@ -1,6 +1,6 @@
-from typing import Any, Optional, TypedDict, Literal
-from typing_extensions import NotRequired
+from typing import Any, Literal, Optional, TypedDict
 
+from typing_extensions import NotRequired
 
 # See: https://ldh.genome.network/docs/ldh/submit.html#content-submission-body
 
@@ -152,3 +152,15 @@ class ClinGenAlleleDefinition(TypedDict):
         "aminoAcidAlleles": NotRequired[list[ClinGenAlleleDefinition]],
     },
 )
+
+ClinGenSubmissionError = TypedDict(
+    "ClinGenSubmissionError",
+    {
+        "description": str,
+        "errorType": str,
+        "hgvs": str,
+        "inputLine": str,
+        "message": str,
+        "position": str,
+    },
+)

tests/lib/clingen/test_services.py

@@ -1,26 +1,26 @@
 # ruff: noqa: E402
 
 import os
-import pytest
-import requests
 from datetime import datetime
-from unittest.mock import patch, MagicMock
+from unittest.mock import MagicMock, patch
 from urllib import parse
 
+import pytest
+import requests
+
 arq = pytest.importorskip("arq")
 cdot = pytest.importorskip("cdot")
 fastapi = pytest.importorskip("fastapi")
 
-from mavedb.lib.clingen.constants import LDH_MAVE_ACCESS_ENDPOINT, GENBOREE_ACCOUNT_NAME, GENBOREE_ACCOUNT_PASSWORD
-from mavedb.lib.utils import batched
+from mavedb.lib.clingen.constants import GENBOREE_ACCOUNT_NAME, GENBOREE_ACCOUNT_PASSWORD, LDH_MAVE_ACCESS_ENDPOINT
 from mavedb.lib.clingen.services import (
     ClinGenAlleleRegistryService,
     ClinGenLdhService,
-    get_clingen_variation,
     clingen_allele_id_from_ldh_variation,
     get_allele_registry_associations,
+    get_clingen_variation,
 )
-
+from mavedb.lib.utils import batched
 from tests.helpers.constants import VALID_CLINGEN_CA_ID
 
 TEST_CLINGEN_URL = "https://pytest.clingen.com"
@@ -365,3 +365,27 @@ def test_get_allele_registry_associations_no_match():
     ]
     result = get_allele_registry_associations(content_submissions, submission_response)
     assert result == {}
+
+
+def test_get_allele_registry_associations_mixed():
+    content_submissions = ["NM_0001:c.1A>G", "NM_0002:c.2T>C", "NM_0003:c.3G>A"]
+    submission_response = [
+        {
+            "@id": "http://reg.test.genome.network/allele/CA123",
+            "genomicAlleles": [{"hgvs": "NM_0001:c.1A>G"}],
+            "transcriptAlleles": [],
+        },
+        {
+            "errorType": "InvalidHGVS",
+            "hgvs": "NM_0002:c.2T>C",
+            "message": "The HGVS string is invalid.",
+        },
+        {
+            "@id": "http://reg.test.genome.network/allele/CA789",
+            "genomicAlleles": [],
+            "transcriptAlleles": [{"hgvs": "NM_0003:c.3G>A"}],
+        },
+    ]
+
+    result = get_allele_registry_associations(content_submissions, submission_response)
+    assert result == {"NM_0001:c.1A>G": "CA123", "NM_0003:c.3G>A": "CA789"}