pyobo 0.10.5__py3-none-any.whl → 0.10.7__py3-none-any.whl

pyobo/sources/hgnc.py

@@ -27,6 +27,7 @@ from pyobo.struct import (
     orthologous,
     transcribes_to,
 )
+from pyobo.struct.typedef import exact_match
 from pyobo.utils.path import ensure_path, prefix_directory_join
 
 __all__ = [
@@ -108,6 +109,28 @@ ENCODINGS = {
     "unknown": "GRP",
 }
 
+SKIP_KEYS = {
+    "date_approved_reserved",
+    "_version_",
+    "uuid",
+    "date_modified",
+    "date_name_changed",
+    "date_symbol_changed",
+    "symbol_report_tag",
+    "location_sortable",
+    "curator_notes",
+    "agr",  # repeat of HGNC ID
+    "gencc",  # repeat of HGNC ID
+    "bioparadigms_slc",  # repeat of symbol
+    "lncrnadb",  # repeat of symbol
+    "gtrnadb",  # repeat of symbol
+    "horde_id",  # repeat of symbol
+    "imgt",  # repeat of symbol
+    "cd",  # symbol
+    "homeodb",  # TODO add to bioregistry, though this is defunct
+    "mamit-trnadb",  # TODO add to bioregistry, though this is defunct
+}
+
 #: A mapping from HGNC's locus_type annotations to sequence ontology identifiers
 LOCUS_TYPE_TO_SO = {
     # protein-coding gene
@@ -190,6 +213,7 @@ class HGNCGetter(Obo):
         transcribes_to,
         orthologous,
         member_of,
+        exact_match,
     ]
     idspaces = IDSPACES
     synonym_typedefs = [
@@ -330,6 +354,12 @@ def get_terms(version: Optional[str] = None, force: bool = False) -> Iterable[Te
             else:
                 tqdm.write(f"unhandled IUPHAR: {iuphar}")
 
+        for lrg_info in entry.pop("lsdb", []):
+            if lrg_info.startswith("LRG_"):
+                lrg_curie = lrg_info.split("|")[0]
+                _, lrg_id = lrg_curie.split("_")
+                term.append_xref(Reference(prefix="lrg", identifier=lrg_id))
+
         for xref_prefix, key in gene_xrefs:
             xref_identifiers = entry.pop(key, None)
             if xref_identifiers is None:
@@ -397,7 +427,8 @@ def get_terms(version: Optional[str] = None, force: bool = False) -> Iterable[Te
         term.set_species(identifier="9606", name="Homo sapiens")
 
         for key in entry:
-            unhandled_entry_keys[key] += 1
+            if key not in SKIP_KEYS:
+                unhandled_entry_keys[key] += 1
         yield term
 
     with open(prefix_directory_join(PREFIX, name="unhandled.json"), "w") as file:

pyobo/sources/mgi.py

@@ -9,6 +9,8 @@ from typing import Iterable
 import pandas as pd
 from tqdm.auto import tqdm
 
+from pyobo.struct.typedef import exact_match
+
 from ..struct import (
     Obo,
     Reference,
@@ -37,7 +39,7 @@ class MGIGetter(Obo):
 
     ontology = PREFIX
     dynamic_version = True
-    typedefs = [from_species, has_gene_product, transcribes_to]
+    typedefs = [from_species, has_gene_product, transcribes_to, exact_match]
 
     def iter_terms(self, force: bool = False) -> Iterable[Term]:
         """Iterate over terms in the ontology."""

pyobo/sources/mirbase.py

@@ -136,6 +136,8 @@ def _process_definitions_lines(
             xref_prefix, xref_identifier, xref_label = map(str.strip, line.split(";"))
             xref_prefix = xref_prefix.lower()
             xref_prefix = xref_mapping.get(xref_prefix, xref_prefix)
+            if xref_prefix == "pictar":
+                continue
             xrefs.append(
                 Reference(prefix=xref_prefix, identifier=xref_identifier, name=xref_label or None)
             )

pyobo/sources/mirbase_family.py

@@ -26,6 +26,7 @@ class MiRBaseFamilyGetter(Obo):
 
     ontology = PREFIX
     bioversions_key = "mirbase"
+    typedefs = [has_member]
 
     def iter_terms(self, force: bool = False) -> Iterable[Term]:
         """Iterate over terms in the ontology."""
@@ -40,7 +41,9 @@ def get_obo(force: bool = False) -> Obo:
 def iter_terms(version: str, force: bool = False) -> Iterable[Term]:
     """Get miRBase family terms."""
     df = get_df(version, force=force)
-    for family_id, name, mirna_id, mirna_name in tqdm(df.values, total=len(df.index)):
+    for family_id, name, mirna_id, mirna_name in tqdm(
+        df.values, total=len(df.index), unit_scale=True, desc="miRBase Family"
+    ):
         term = Term(
             reference=Reference(prefix=PREFIX, identifier=family_id, name=name),
         )
@@ -65,4 +68,4 @@ def get_df(version: str, force: bool = False) -> pd.DataFrame:
 
 
 if __name__ == "__main__":
-    get_obo().write_default(use_tqdm=True)
+    get_obo().write_default(use_tqdm=True, write_obo=True, force=True)

pyobo/sources/npass.py

@@ -72,12 +72,14 @@ def iter_terms(version: str, force: bool = False) -> Iterable[Term]:
 
         # TODO check that the first is always the parent compound?
         if pd.notna(pubchem_compound_ids):
-            pubchem_compound_ids = pubchem_compound_ids.split(";")
+            pubchem_compound_ids = [
+                yy.strip() for xx in pubchem_compound_ids.split(";") for yy in xx.strip().split(",")
+            ]
             if len(pubchem_compound_ids) > 1:
                 logger.debug("multiple cids for %s: %s", identifier, pubchem_compound_ids)
             for pubchem_compound_id in pubchem_compound_ids:
                 term.append_xref(
-                    Reference(prefix="pubchem.compound", identifier=pubchem_compound_id)
+                    Reference(prefix="pubchem.compound", identifier=pubchem_compound_id.strip())
                 )
 
         for synonym in [iupac]:

pyobo/sources/pombase.py

@@ -29,7 +29,7 @@ class PomBaseGetter(Obo):
     """An ontology representation of PomBase's fission yeast gene nomenclature."""
 
     ontology = bioversions_key = PREFIX
-    typedefs = [from_species, has_gene_product]
+    typedefs = [from_species, has_gene_product, orthologous]
 
     def iter_terms(self, force: bool = False) -> Iterable[Term]:
         """Iterate over terms in the ontology."""

pyobo/sources/ror.py

@@ -0,0 +1,163 @@
+"""Convert the Research Organization Registry (ROR) into an ontology."""
+
+import json
+import zipfile
+from typing import Iterable
+
+import bioregistry
+import zenodo_client
+from tqdm.auto import tqdm
+
+from pyobo.struct import Obo, Reference, Term, TypeDef
+from pyobo.struct.struct import acronym
+
+PREFIX = "ror"
+ROR_ZENODO_RECORD_ID = "10086202"
+
+# Constants
+ORG_CLASS = Reference(prefix="OBI", identifier="0000245")
+LOCATED_IN = Reference(prefix="RO", identifier="0001025")
+PART_OF = Reference(prefix="BFO", identifier="0000050")
+HAS_PART = Reference(prefix="BFO", identifier="0000051")
+SUCCESSOR = Reference(prefix="BFO", identifier="0000063")
+PREDECESSOR = Reference(prefix="BFO", identifier="0000062")
+
+RMAP = {
+    "Related": TypeDef.from_triple("rdfs", "seeAlso"),
+    "Child": TypeDef(HAS_PART),
+    "Parent": TypeDef(PART_OF),
+    "Predecessor": TypeDef(PREDECESSOR),
+    "Successor": TypeDef(SUCCESSOR),
+    "Located in": TypeDef(LOCATED_IN),
+}
+NAME_REMAPPING = {
+    "'s-Hertogenbosch": "Den Bosch",  # SMH Netherlands, why u gotta be like this
+    "'s Heeren Loo": "s Heeren Loo",
+    "'s-Heerenberg": "s-Heerenberg",
+    "Institut Virion\\Serion": "Institut Virion/Serion",
+    "Hematology\\Oncology Clinic": "Hematology/Oncology Clinic",
+}
+
+
+class RORGetter(Obo):
+    """An ontology representation of the ROR."""
+
+    ontology = bioregistry_key = PREFIX
+    typedefs = list(RMAP.values())
+    synonym_typedefs = [acronym]
+    idspaces = {
+        "ror": "https://ror.org/",
+        "geonames": "https://www.geonames.org/",
+        "envo": "http://purl.obolibrary.org/obo/ENVO_",
+        "bfo": "http://purl.obolibrary.org/obo/BFO_",
+        "ro": "http://purl.obolibrary.org/obo/RO_",
+        "obi": "http://purl.obolibrary.org/obo/OBI_",
+        "omo": "http://purl.obolibrary.org/obo/OMO_",
+        "rdfs": "http://www.w3.org/2000/01/rdf-schema#",
+    }
+
+    def __post_init__(self):  # noqa: D105
+        self.data_version, _url, _path = _get_info()
+        super().__post_init__()
+
+    def iter_terms(self, force: bool = False) -> Iterable[Term]:
+        """Iterate over terms in the ontology."""
+        return iterate_ror_terms(force=force)
+
+
+def iterate_ror_terms(*, force: bool = False) -> Iterable[Term]:
+    """Iterate over terms in ROR."""
+    version, source_uri, records = get_latest(force=force)
+    unhandled_xref_prefixes = set()
+    for record in tqdm(records, unit_scale=True, unit="record", desc=PREFIX):
+        identifier = record["id"].removeprefix("https://ror.org/")
+        name = record["name"]
+        name = NAME_REMAPPING.get(name, name)
+
+        term = Term(
+            reference=Reference(prefix=PREFIX, identifier=identifier, name=name), type="Instance"
+        )
+        term.append_parent(ORG_CLASS)
+
+        if name.startswith("The "):
+            term.append_synonym(name.removeprefix("The "))
+
+        for relationship in record.get("relationships", []):
+            target_id = relationship["id"].removeprefix("https://ror.org/")
+            term.append_relationship(
+                RMAP[relationship["type"]], Reference(prefix=PREFIX, identifier=target_id)
+            )
+
+        term.is_obsolete = record.get("status") != "active"
+
+        for address in record.get("addresses", []):
+            city = address.get("geonames_city")
+            if not city:
+                continue
+            term.append_relationship(
+                RMAP["Located in"], Reference(prefix="geonames", identifier=str(city["id"]))
+            )
+
+        for label in record.get("labels", []):
+            label = label["label"]  # there's a language availabel in this dict too
+            term.append_synonym(label)
+            if label.startswith("The "):
+                term.append_synonym(label.removeprefix("The "))
+
+        for synonym in record.get("aliases", []):
+            term.append_synonym(synonym)
+            if synonym.startswith("The "):
+                term.append_synonym(synonym.removeprefix("The "))
+
+        for acronym_synonym in record.get("acronyms", []):
+            term.append_synonym(acronym_synonym, type=acronym)
+
+        for prefix, xref_data in record.get("external_ids", {}).items():
+            if prefix == "OrgRef":
+                # OrgRef refers to wikipedia page id, see
+                # https://stackoverflow.com/questions/6168020/what-is-wikipedia-pageid-how-to-change-it-into-real-page-url
+                continue
+            norm_prefix = bioregistry.normalize_prefix(prefix)
+            if norm_prefix is None:
+                if prefix not in unhandled_xref_prefixes:
+                    tqdm.write(f"Unhandled prefix: {prefix} in {name} ({term.curie}). Values:")
+                    for xref_id in xref_data["all"]:
+                        tqdm.write(f"- {xref_id}")
+                    unhandled_xref_prefixes.add(prefix)
+                continue
+
+            identifiers = xref_data["all"]
+            if isinstance(identifiers, str):
+                identifiers = [identifiers]
+            for xref_id in identifiers:
+                term.append_xref(Reference(prefix=norm_prefix, identifier=xref_id.replace(" ", "")))
+
+        yield term
+
+
+def _get_info(*, force: bool = False):
+    client = zenodo_client.Zenodo()
+    latest_record_id = client.get_latest_record(ROR_ZENODO_RECORD_ID)
+    response = client.get_record(latest_record_id)
+    response_json = response.json()
+    version = response_json["metadata"]["version"].lstrip("v")
+    file_record = response_json["files"][0]
+    name = file_record["key"]
+    url = file_record["links"]["self"]
+    path = client.download(latest_record_id, name=name, force=force)
+    return version, url, path
+
+
+def get_latest(*, force: bool = False):
+    """Get the latest ROR metadata and records."""
+    version, url, path = _get_info(force=force)
+    with zipfile.ZipFile(path) as zf:
+        for zip_info in zf.filelist:
+            if zip_info.filename.endswith(".json"):
+                with zf.open(zip_info) as file:
+                    return version, url, json.load(file)
+    raise FileNotFoundError
+
+
+if __name__ == "__main__":
+    RORGetter().write_default(write_obo=True, force=True)

pyobo/sources/sgd.py

@@ -5,7 +5,7 @@
 from typing import Iterable
 from urllib.parse import unquote_plus
 
-from ..struct import Obo, Reference, Synonym, SynonymTypeDef, Term, from_species
+from ..struct import Obo, Reference, Synonym, Term, from_species
 from ..utils.path import ensure_tar_df
 
 __all__ = [
@@ -21,15 +21,12 @@ URL = (
 )
 INNER_PATH = "S288C_reference_genome_R64-2-1_20150113/saccharomyces_cerevisiae_R64-2-1_20150113.gff"
 
-alias_type = SynonymTypeDef.from_text("alias")
-
 
 class SGDGetter(Obo):
     """An ontology representation of SGD's yeast gene nomenclature."""
 
     bioversions_key = ontology = PREFIX
     typedefs = [from_species]
-    synonym_typedefs = [alias_type]
 
     def iter_terms(self, force: bool = False) -> Iterable[Term]:
         """Iterate over terms for SGD."""
@@ -68,7 +65,7 @@ def get_terms(ontology: Obo, force: bool = False) -> Iterable[Term]:
         aliases = d.get("Alias")
         if aliases:
             for alias in aliases.split(","):
-                synonyms.append(Synonym(name=unquote_plus(alias), type=alias_type))
+                synonyms.append(Synonym(name=unquote_plus(alias)))
 
         term = Term(
             reference=Reference(prefix=PREFIX, identifier=identifier, name=name),

pyobo/sources/slm.py

@@ -7,8 +7,9 @@ from typing import Iterable
 import pandas as pd
 from tqdm.auto import tqdm
 
-from pyobo import Obo, Reference, SynonymTypeDef, Term
-from pyobo.struct.typedef import has_inchi, has_smiles
+from pyobo import Obo, Reference, Term
+from pyobo.struct.struct import abbreviation as abbreviation_typedef
+from pyobo.struct.typedef import exact_match, has_inchi, has_smiles
 from pyobo.utils.path import ensure_df
 
 __all__ = [
@@ -38,14 +39,13 @@ COLUMNS = [
     "PMID",
 ]
 
-abreviation_type = SynonymTypeDef.from_text("abbreviation")
-
 
 class SLMGetter(Obo):
     """An ontology representation of SwissLipid's lipid nomenclature."""
 
     ontology = bioversions_key = PREFIX
-    synonym_typedefs = [abreviation_type]
+    typedefs = [exact_match]
+    synonym_typedefs = [abbreviation_typedef]
 
     def iter_terms(self, force: bool = False) -> Iterable[Term]:
         """Iterate over terms in the ontology."""
@@ -94,7 +94,7 @@ def iter_terms(version: str, force: bool = False):
         if pd.notna(level):
             term.append_property("level", level)
         if pd.notna(abbreviation):
-            term.append_synonym(abbreviation, type=abreviation_type)
+            term.append_synonym(abbreviation, type=abbreviation_typedef)
         if pd.notna(synonyms):
             for synonym in synonyms.split("|"):
                 term.append_synonym(synonym.strip())

pyobo/sources/umls/get_synonym_types.py

@@ -0,0 +1,36 @@
+"""Utilities for UMLS synonyms."""
+
+from pathlib import Path
+from typing import Mapping
+
+import requests
+from bs4 import BeautifulSoup
+
+from pyobo.utils.io import open_map_tsv, write_map_tsv
+
+__all__ = ["get_umls_synonyms"]
+
+HERE = Path(__file__).parent.resolve()
+SYNONYM_TYPE_PATH = HERE.joinpath("synonym_types.tsv")
+
+ABBREVIATIONS_URL = "https://www.nlm.nih.gov/research/umls/knowledge_sources/metathesaurus/release/abbreviations.html"
+
+
+def get_umls_synonyms(*, refresh: bool = False) -> Mapping[str, str]:
+    """Get all synonyms."""
+    if SYNONYM_TYPE_PATH.is_file() and not refresh:
+        return open_map_tsv(SYNONYM_TYPE_PATH)
+    res = requests.get(ABBREVIATIONS_URL, timeout=5)
+    soup = BeautifulSoup(res.text, features="html.parser")
+    table = soup.find(id="mrdoc_TTY")
+    body = table.find("tbody")
+    rv = {}
+    for row in body.find_all("tr"):
+        left, right = row.find_all("td")
+        rv[left.text.strip()] = right.text.strip()
+    write_map_tsv(path=SYNONYM_TYPE_PATH, rv=rv, header=["key", "name"])
+    return rv
+
+
+if __name__ == "__main__":
+    get_umls_synonyms(refresh=True)