feat(uniprot): support Ensembl protein accessions in UniProt ID mapping

src/mavedb/lib/uniprot/utils.py

@@ -1,11 +1,11 @@
 from typing import Literal, Union
 
-from mavedb.lib.validation.identifier import validate_refseq_identifier
+from mavedb.lib.validation.identifier import validate_ensembl_identifier, validate_refseq_identifier
 
 
 def infer_db_name_from_sequence_accession(
     sequence_accession: str,
-) -> Union[Literal["RefSeq_Nucleotide", "RefSeq_Protein"]]:
+) -> Union[Literal["RefSeq_Nucleotide", "RefSeq_Protein", "Ensembl_Protein"]]:
     """
     Infers the database name from a sequence accession.
 
@@ -15,11 +15,19 @@ def infer_db_name_from_sequence_accession(
     Returns:
         str: The inferred database name.
     """
+    # Ensembl protein accessions must be handled before validate_refseq_identifier,
+    # which raises on any non-RefSeq accession.
+    if sequence_accession.startswith("ENSP"):
+        validate_ensembl_identifier(sequence_accession)
+        return "Ensembl_Protein"
+
     validate_refseq_identifier(sequence_accession)
 
     if sequence_accession.startswith("NM_"):
         return "RefSeq_Nucleotide"
     if sequence_accession.startswith("NP_"):
         return "RefSeq_Protein"
 
-    raise NotImplementedError("Only RefSeq NM and NP identifiers are currently supported for inference.")
+    raise NotImplementedError(
+        "Only RefSeq NM/NP and Ensembl protein (ENSP) identifiers are currently supported for inference."
+    )

src/mavedb/lib/validation/identifier.py

@@ -1,3 +1,5 @@
+import re
+
 import idutils
 
 from mavedb.lib.validation.constants.identifier import valid_dbnames
@@ -52,6 +54,9 @@ def validate_ensembl_identifier(identifier: str):
     """
     Validates whether the identifier is a valid Ensembl identifier.
 
+    Ensembl stable IDs may carry a version suffix (e.g. ``ENSP00000369497.3``), which
+    ``idutils.is_ensembl`` does not accept; strip it before validating the stable ID.
+
     Parameters
     __________
     identifier : str
@@ -62,7 +67,8 @@ def validate_ensembl_identifier(identifier: str):
     ValidationError
         If the identifier is not a valid Ensembl identifier.
     """
-    if not idutils.is_ensembl(identifier):
+    base_identifier = re.sub(r"\.\d+$", "", identifier)
+    if not idutils.is_ensembl(base_identifier):
         raise ValidationError(f"'{identifier}' is not a valid Ensembl accession.")
 
 

tests/lib/uniprot/test_utils.py

@@ -21,6 +21,14 @@ def test_infer_db_name_from_sequence_accession_np():
     assert result == "RefSeq_Protein"
 
 
+# Both versioned and unversioned ENSP accessions must route to Ensembl_Protein; the versioned case
+# guards that validate_ensembl_identifier's version handling continues to accept it.
+@pytest.mark.parametrize("ensembl_protein_accession", ["ENSP00000418960", "ENSP00000418960.3"])
+def test_infer_db_name_from_sequence_accession_ensp(ensembl_protein_accession):
+    result = infer_db_name_from_sequence_accession(ensembl_protein_accession)
+    assert result == "Ensembl_Protein"
+
+
 @pytest.mark.parametrize("invalid_accession", ["XP_000000", VALID_CHR_ACCESSION])
 def test_infer_db_name_from_sequence_accession_invalid(invalid_accession):
     with pytest.raises(NotImplementedError):

tests/validation/test_identifier.py

@@ -63,6 +63,10 @@ def test_ve_invalid_list(self):
     def test_passes_valid_id(self):
         validate_ensembl_identifier("ENSG00000143384")
 
+    def test_passes_valid_versioned_id(self):
+        validate_ensembl_identifier("ENSP00000369497.3")
+        validate_ensembl_list(["ENSG00000139618.15", "ENST00000380152.8"])
+
 
 class TestRefSeqValidators(TestCase):
     """