Add UMLS STY

src/pyobo/sources/__init__.py

@@ -50,7 +50,7 @@
 from .sgd import SGDGetter
 from .signor import SignorGetter
 from .slm import SLMGetter
-from .umls import UMLSGetter
+from .umls import UMLSGetter, UMLSSTyGetter
 from .unimod import UnimodGetter
 from .uniprot import UniProtGetter, UniProtPtmGetter
 from .wikipathways import WikiPathwaysGetter
@@ -125,6 +125,7 @@
     "SLMGetter",
     "SignorGetter",
     "UMLSGetter",
+    "UMLSSTyGetter",
     "UniProtGetter",
     "UniProtPtmGetter",
     "UnimodGetter",

src/pyobo/sources/umls/__init__.py

@@ -1,7 +1,9 @@
 """Converter for UMLS."""
 
+from .sty import UMLSSTyGetter
 from .umls import UMLSGetter
 
 __all__ = [
     "UMLSGetter",
+    "UMLSSTyGetter",
 ]

src/pyobo/sources/umls/sty.py

@@ -0,0 +1,57 @@
+"""Converter for UMLS Semantic Types."""
+
+from collections.abc import Iterable
+
+from pyobo import Obo, Reference, Term, default_reference
+from pyobo.struct.typedef import has_category
+from pyobo.utils.path import ensure_df
+
+__all__ = [
+    "UMLSSTyGetter",
+]
+
+PREFIX = "sty"
+
+URL = "https://www.nlm.nih.gov/research/umls/knowledge_sources/semantic_network/SemGroups.txt"
+
+
+class UMLSSTyGetter(Obo):
+    """An ontology representation of UMLS Semantic Types."""
+
+    ontology = PREFIX
+    bioversions_key = "umls"
+    typedefs = [has_category]
+
+    def iter_terms(self, force: bool = False) -> Iterable[Term]:
+        """Iterate over terms in the ontology."""
+        return iter_terms(version=self._version_or_raise)
+
+
+COLUMNS = [
+    "group",
+    "group_label",
+    "sty_id",
+    "sty_name",
+]
+
+
+def iter_terms(version: str) -> Iterable[Term]:
+    """Iterate over UMLS terms."""
+    df = ensure_df(PREFIX, url=URL, version=version, sep="|", header=None, names=COLUMNS)
+
+    extras = {
+        group: Term(
+            reference=default_reference(PREFIX, group, name=group_label),
+        )
+        for group, group_label in df[["group", "group_label"]].drop_duplicates().values
+    }
+    yield from extras.values()
+
+    for group, _group_label, sty_id, sty_name in df.values:
+        term = Term(reference=Reference(prefix="sty", identifier=sty_id, name=sty_name))
+        term.append_parent(extras[group])
+        yield term
+
+
+if __name__ == "__main__":
+    UMLSSTyGetter.cli()