Load molecular structures properly

fastdatascience · Aug 8, 2024 · b00d195 · b00d195
1 parent 7ca9328
commit b00d195
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 1 deletion.
diff --git a/harvesting_data_from_source/05_combine_data_sources.py b/harvesting_data_from_source/05_combine_data_sources.py
@@ -50,7 +50,11 @@
 
 def add_canonical(canonical: str, data: dict):
     canonical_norm = canonical.lower().strip()
-    data["name"] = canonical
+    if canonical_norm in drug_variant_to_canonical and canonical_norm not in drug_variant_to_canonical[canonical_norm]:
+        print(f"Adding canonical {canonical_norm} but it already maps to {drug_variant_to_canonical[canonical_norm]}")
+        canonical_norm = drug_variant_to_canonical[canonical_norm][0]
+    elif canonical_norm not in drug_variant_to_canonical:
+        data["name"] = canonical
     if canonical_norm not in drug_canonical_to_data:
         drug_canonical_to_data[canonical_norm] = data
     else:

diff --git a/src/drug_named_entity_recognition/drug_ner_dictionary.pkl.bz2 b/src/drug_named_entity_recognition/drug_ner_dictionary.pkl.bz2