pyobo 0.12.3__py3-none-any.whl → 0.12.5__py3-none-any.whl

pyobo/__init__.py

@@ -48,6 +48,8 @@ from .api import (
     get_sssom_df,
     get_subhierarchy,
     get_synonyms,
+    get_text_embedding,
+    get_text_embedding_similarity,
     get_typedef_df,
     get_xref,
     get_xrefs,
@@ -70,6 +72,7 @@ from .struct import (
     SynonymTypeDef,
     Term,
     TypeDef,
+    build_ontology,
     default_reference,
 )
 from .struct.obo import from_obo_path, from_obonet
@@ -84,6 +87,7 @@ __all__ = [
     "SynonymTypeDef",
     "Term",
     "TypeDef",
+    "build_ontology",
     "default_reference",
     "ensure_path",
     "from_obo_path",
@@ -137,6 +141,8 @@ __all__ = [
     "get_sssom_df",
     "get_subhierarchy",
     "get_synonyms",
+    "get_text_embedding",
+    "get_text_embedding_similarity",
     "get_typedef_df",
     "get_version",
     "get_xref",

pyobo/api/__init__.py

@@ -8,6 +8,7 @@ from .alts import (
 )
 from .combine import get_literal_mappings_subset
 from .edges import get_edges, get_edges_df, get_graph
+from .embedding import get_text_embedding, get_text_embedding_similarity
 from .hierarchy import (
     get_ancestors,
     get_children,
@@ -116,6 +117,8 @@ __all__ = [
     "get_sssom_df",
     "get_subhierarchy",
     "get_synonyms",
+    "get_text_embedding",
+    "get_text_embedding_similarity",
     "get_typedef_df",
     "get_version",
     "get_xref",

pyobo/api/embedding.py

@@ -0,0 +1,118 @@
+"""Embeddings for entities."""
+
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
+import curies
+import numpy as np
+
+from pyobo.api.names import get_definition, get_name
+
+if TYPE_CHECKING:
+    import sentence_transformers
+
+__all__ = [
+    "get_text_embedding",
+    "get_text_embedding_model",
+    "get_text_embedding_similarity",
+]
+
+
+def get_text_embedding_model() -> sentence_transformers.SentenceTransformer:
+    """Get the default text embedding model."""
+    from sentence_transformers import SentenceTransformer
+
+    model = SentenceTransformer("all-MiniLM-L6-v2")
+    return model
+
+
+def _get_text(
+    reference: str | curies.Reference | curies.ReferenceTuple,
+) -> str | None:
+    name = get_name(reference)
+    if name is None:
+        return None
+    description = get_definition(reference)
+    if description:
+        name += " " + description
+    return name
+
+
+def get_text_embedding(
+    reference: str | curies.Reference | curies.ReferenceTuple,
+    *,
+    model: sentence_transformers.SentenceTransformer | None = None,
+) -> np.ndarray | None:
+    """Get a text embedding for an entity, or return none if no text is available.
+
+    :param reference: A reference, either as a string or Reference object
+    :param model: A sentence transformer model. Defaults to ``all-MiniLM-L6-v2`` if not given.
+    :return: A 1D numpy float array of embeddings from :class:`sentence_transformers`
+
+    .. code-block:: python
+
+        import pyobo
+
+        embedding = pyobo.get_text_embedding("GO:0000001")
+        # [-5.68335280e-02  7.96175096e-03 -3.36112119e-02  2.34440481e-03 ... ]
+
+    If you want to do multiple operations, load up the model for reuse
+
+    .. code-block:: python
+
+        import pyobo
+        from pyobo.api.embedding import get_text_embedding_model
+
+        model = get_text_embedding_model()
+        embedding = pyobo.get_text_embedding("GO:0000001", model=model)
+        # [-5.68335280e-02  7.96175096e-03 -3.36112119e-02  2.34440481e-03 ... ]
+    """
+    text = _get_text(reference)
+    if text is None:
+        return None
+    if model is None:
+        model = get_text_embedding_model()
+    res = model.encode([text])
+    return res[0]
+
+
+def get_text_embedding_similarity(
+    reference_1: str | curies.Reference | curies.ReferenceTuple,
+    reference_2: str | curies.Reference | curies.ReferenceTuple,
+    *,
+    model: sentence_transformers.SentenceTransformer | None = None,
+) -> float | None:
+    """Get the pairwise similarity.
+
+    :param reference_1: A reference, given as a string or Reference object
+    :param reference_2: A second reference
+    :param model: A sentence transformer model. Defaults to ``all-MiniLM-L6-v2`` if not given.
+    :returns:
+        A floating point similarity, if text is available for both references, otherwise none
+
+    .. code-block:: python
+
+        import pyobo
+
+        similarity = pyobo.get_text_embedding_similarity("GO:0000001", "GO:0000004")
+        # 0.24702128767967224
+
+    If you want to do multiple operations, load up the model for reuse
+
+    .. code-block:: python
+
+        import pyobo
+        from pyobo.api.embedding import get_text_embedding_model
+
+        model = get_text_embedding_model()
+        similarity = pyobo.get_text_embedding_similarity("GO:0000001", "GO:0000004", model=model)
+        # 0.24702128767967224
+    """
+    if model is None:
+        model = get_text_embedding_model()
+    e1 = get_text_embedding(reference_1, model=model)
+    e2 = get_text_embedding(reference_2, model=model)
+    if e1 is None or e2 is None:
+        return None
+    return model.similarity(e1, e2)[0][0].item()

pyobo/api/names.py

@@ -166,7 +166,14 @@ def get_references(prefix: str, **kwargs: Unpack[GetOntologyKwargs]) -> set[Refe
         ontology = get_ontology(prefix, **kwargs)
         return sorted(ontology.iterate_references())
 
-    return set(_get_references())
+    try:
+        return set(_get_references())
+    except NoBuildError:
+        logger.debug("[%s] no build", prefix)
+        return set()
+    except (Exception, subprocess.CalledProcessError) as e:
+        logger.exception("[%s v%s] could not load: %s", prefix, version, e)
+        return set()
 
 
 @lru_cache

pyobo/api/utils.py

@@ -19,7 +19,6 @@ __all__ = [
     "VersionError",
     "get_version",
     "get_version_pins",
-    "safe_get_version",
 ]
 
 logger = logging.getLogger(__name__)
@@ -84,15 +83,6 @@ def get_version_from_kwargs(prefix: str, kwargs: GetOntologyKwargs) -> str | Non
     return get_version(prefix, strict=False)
 
 
-def safe_get_version(prefix: str) -> str:
-    """Get the version."""
-    # FIXME replace with get_version(prefix, strict=True)
-    v = get_version(prefix)
-    if v is None:
-        raise ValueError
-    return v
-
-
 @lru_cache(1)
 def get_version_pins() -> dict[str, str]:
     """Retrieve user-defined resource version pins.

pyobo/cli/cli.py

@@ -90,12 +90,7 @@ def _has_no_download(prefix: str) -> bool:
 @lru_cache(maxsize=1)
 def _no_download() -> set[str]:
     """Get the list of prefixes not available as OBO."""
-    return {
-        prefix
-        for prefix in bioregistry.read_registry()
-        if bioregistry.get_obo_download(prefix) is None
-        and bioregistry.get_owl_download(prefix) is None
-    }
+    return {resource.prefix for resource in bioregistry.resources() if not resource.has_download()}
 
 
 main.add_command(lookup)

pyobo/constants.py

@@ -4,7 +4,11 @@ from __future__ import annotations
 
 import logging
 import re
+from collections.abc import Callable
+from pathlib import Path
+from typing import Literal, NamedTuple, TypeAlias
 
+import bioregistry
 import pystow
 from typing_extensions import NotRequired, TypedDict
 
@@ -188,6 +192,8 @@ class IterHelperHelperDict(SlimGetOntologyKwargs):
     skip_set: set[str] | None
 
 
+OntologyFormat: TypeAlias = Literal["obo", "owl", "json", "rdf"]
+
 #: from table 2 of the Functional OWL syntax definition
 #: at https://www.w3.org/TR/owl2-syntax/#IRIs
 DEFAULT_PREFIX_MAP = {
@@ -196,3 +202,20 @@ DEFAULT_PREFIX_MAP = {
     "xsd": "http://www.w3.org/2001/XMLSchema#",
     "owl": "http://www.w3.org/2002/07/owl#",
 }
+
+
+class OntologyPathPack(NamedTuple):
+    """A format and path tuple."""
+
+    format: OntologyFormat
+    path: Path
+
+
+#: Functions that get ontology files. Order matters in this list,
+#: since order implicitly defines priority
+ONTOLOGY_GETTERS: list[tuple[OntologyFormat, Callable[[str], str | None]]] = [
+    ("obo", bioregistry.get_obo_download),
+    ("owl", bioregistry.get_owl_download),
+    ("json", bioregistry.get_json_download),
+    ("rdf", bioregistry.get_rdf_download),
+]

pyobo/getters.py

@@ -17,9 +17,11 @@ from pathlib import Path
 from textwrap import indent
 from typing import Any, TypeVar
 
+import bioontologies.robot
 import bioregistry
 import click
 import pystow.utils
+import requests.exceptions
 from tabulate import tabulate
 from tqdm.auto import tqdm
 from typing_extensions import Unpack
@@ -27,8 +29,10 @@ from typing_extensions import Unpack
 from .constants import (
     BUILD_SUBDIRECTORY_NAME,
     DATABASE_DIRECTORY,
+    ONTOLOGY_GETTERS,
     GetOntologyKwargs,
     IterHelperHelperDict,
+    OntologyPathPack,
     SlimGetOntologyKwargs,
 )
 from .identifier_utils import ParseError, wrap_norm_prefix
@@ -36,6 +40,7 @@ from .plugins import has_nomenclature_plugin, run_nomenclature_plugin
 from .struct import Obo
 from .struct.obo import from_obo_path, from_obonet
 from .utils.io import safe_open_writer
+from .utils.misc import _get_version_from_artifact
 from .utils.path import ensure_path, prefix_directory_join
 from .version import get_git_hash, get_version
 
@@ -56,7 +61,7 @@ class UnhandledFormatError(NoBuildError):
 
 
 #: The following prefixes can not be loaded through ROBOT without
-#: turning off integrity checks
+#: turning off integrity checks. This used to be part of _convert_to_obo
 REQUIRES_NO_ROBOT_CHECK = {
     "clo",
     "vo",
@@ -64,9 +69,18 @@ REQUIRES_NO_ROBOT_CHECK = {
     "orphanet",
     "foodon",
     "caloha",
+    # "aeon",
 }
 
 
+def _convert_to_obo(path: Path) -> Path:
+    import bioontologies.robot
+
+    _converted_obo_path = path.with_suffix(".obo")
+    bioontologies.robot.convert(path, _converted_obo_path, check=False)
+    return _converted_obo_path
+
+
 @wrap_norm_prefix
 def get_ontology(
     prefix: str,
@@ -114,10 +128,20 @@ def get_ontology(
     """
     if force:
         force_process = True
+    if has_nomenclature_plugin(prefix):
+        obo = run_nomenclature_plugin(prefix, version=version)
+        if cache:
+            logger.debug("[%s] caching nomenclature plugin", prefix)
+            obo.write_default(force=force_process)
+        return obo
+
     if prefix == "uberon":
         logger.info("UBERON has so much garbage in it that defaulting to non-strict parsing")
         strict = False
 
+    if version is None:
+        version = _get_version_from_artifact(prefix)
+
     if force_process:
         obonet_json_gz_path = None
     elif not cache:
@@ -146,26 +170,14 @@ def get_ontology(
         else:
             logger.debug("[%s] no obonet cache found at %s", prefix, obonet_json_gz_path)
 
-    if has_nomenclature_plugin(prefix):
-        obo = run_nomenclature_plugin(prefix, version=version)
-        if cache:
-            logger.debug("[%s] caching nomenclature plugin", prefix)
-            obo.write_default(force=force_process)
-        return obo
-
-    ontology_format, path = _ensure_ontology_path(prefix, force=force, version=version)
-    if path is None:
+    path_pack = _ensure_ontology_path(prefix, force=force, version=version)
+    if path_pack is None:
         raise NoBuildError(prefix)
-    elif ontology_format == "obo":
+    ontology_format, path = path_pack
+    if ontology_format == "obo":
         pass  # all gucci
-    elif ontology_format == "owl":
-        import bioontologies.robot
-
-        _converted_obo_path = path.with_suffix(".obo")
-        if prefix in REQUIRES_NO_ROBOT_CHECK:
-            robot_check = False
-        bioontologies.robot.convert(path, _converted_obo_path, check=robot_check)
-        path = _converted_obo_path
+    elif ontology_format in {"owl", "rdf"}:
+        path = _convert_to_obo(path)
     elif ontology_format == "json":
         from .struct.obograph import read_obograph
 
@@ -191,21 +203,21 @@ def get_ontology(
 
 
 def _ensure_ontology_path(
-    prefix: str, force: bool, version: str | None
-) -> tuple[str, Path] | tuple[None, None]:
-    for ontology_format, url in [
-        ("obo", bioregistry.get_obo_download(prefix)),
-        ("owl", bioregistry.get_owl_download(prefix)),
-        ("json", bioregistry.get_json_download(prefix)),
-    ]:
-        if url is not None:
-            try:
-                path = ensure_path(prefix, url=url, force=force, version=version)
-            except (urllib.error.HTTPError, pystow.utils.DownloadError):
-                continue
-            else:
-                return ontology_format, path
-    return None, None
+    prefix: str, *, force: bool, version: str | None
+) -> OntologyPathPack | None:
+    for ontology_format, getter in ONTOLOGY_GETTERS:
+        url = getter(prefix)
+        if url is None:
+            continue
+        try:
+            path = ensure_path(prefix, url=url, force=force, version=version)
+        except (urllib.error.HTTPError, pystow.utils.DownloadError):
+            continue
+        except pystow.utils.UnexpectedDirectoryError:
+            continue  # TODO report more info about the URL and the name it tried to make
+        else:
+            return OntologyPathPack(ontology_format, path)
+    return None
 
 
 SKIP = {
@@ -234,6 +246,7 @@ SKIP = {
     "gwascentral.phenotype": "website is down? or API changed?",  # FIXME
     "gwascentral.study": "website is down? or API changed?",  # FIXME
     "snomedct": "dead source",
+    "ero": "dead",
 }
 
 X = TypeVar("X")
@@ -349,6 +362,10 @@ def iter_helper_helper(
             logger.warning("[%s] unable to download - %s", prefix, e.reason)
             if strict and not bioregistry.is_deprecated(prefix):
                 raise
+        except requests.exceptions.ConnectTimeout as e:
+            logger.warning("[%s] unable to download - %s", prefix, e)
+            if strict and not bioregistry.is_deprecated(prefix):
+                raise
         except ParseError as e:
             if not e.node:
                 logger.warning("[%s] %s", prefix, e)
@@ -360,7 +377,7 @@ def iter_helper_helper(
             if "DrugBank" not in str(e):
                 raise
             logger.warning("[drugbank] invalid credentials")
-        except subprocess.CalledProcessError:
+        except (subprocess.CalledProcessError, bioontologies.robot.ROBOTError):
             logger.warning("[%s] ROBOT was unable to convert OWL to OBO", prefix)
         except ValueError as e:
             if _is_xml(e):

pyobo/sources/__init__.py

@@ -6,7 +6,13 @@ from .antibodyregistry import AntibodyRegistryGetter
 from .bigg import BiGGCompartmentGetter, BiGGMetaboliteGetter, BiGGModelGetter, BiGGReactionGetter
 from .ccle import CCLEGetter
 from .cgnc import CGNCGetter
-from .chembl import ChEMBLCompoundGetter, ChEMBLTargetGetter
+from .chembl import (
+    ChEMBLCellGetter,
+    ChEMBLCompoundGetter,
+    ChEMBLMechanismGetter,
+    ChEMBLTargetGetter,
+    ChEMBLTissueGetter,
+)
 from .civic_gene import CIVICGeneGetter
 from .clinicaltrials import ClinicalTrialsGetter
 from .complexportal import ComplexPortalGetter
@@ -26,6 +32,7 @@ from .geonames import GeonamesFeatureGetter, GeonamesGetter
 from .gtdb import GTDBGetter
 from .gwascentral import GWASCentralPhenotypeGetter, GWASCentralStudyGetter
 from .hgnc import HGNCGetter, HGNCGroupGetter
+from .iana_media_type import IANAGetter
 from .icd import ICD10Getter, ICD11Getter
 from .intact import IntactGetter
 from .interpro import InterProGetter
@@ -60,6 +67,7 @@ from .selventa import SCHEMGetter, SCOMPGetter, SDISGetter, SFAMGetter
 from .sgd import SGDGetter
 from .signor import SignorGetter
 from .slm import SLMGetter
+from .spdx import SPDXLicenseGetter
 from .umls import UMLSGetter, UMLSSTyGetter
 from .unimod import UnimodGetter
 from .uniprot import UniProtGetter, UniProtPtmGetter
@@ -79,8 +87,11 @@ __all__ = [
     "CONSOGetter",
     "CPTGetter",
     "CVXGetter",
+    "ChEMBLCellGetter",
     "ChEMBLCompoundGetter",
+    "ChEMBLMechanismGetter",
     "ChEMBLTargetGetter",
+    "ChEMBLTissueGetter",
     "ClinicalTrialsGetter",
     "ComplexPortalGetter",
     "CreditGetter",
@@ -100,6 +111,7 @@ __all__ = [
     "GeonamesGetter",
     "HGNCGetter",
     "HGNCGroupGetter",
+    "IANAGetter",
     "ICD10Getter",
     "ICD11Getter",
     "ITISGetter",
@@ -142,6 +154,7 @@ __all__ = [
     "SFAMGetter",
     "SGDGetter",
     "SLMGetter",
+    "SPDXLicenseGetter",
     "SignorGetter",
     "UMLSGetter",
     "UMLSSTyGetter",

pyobo/sources/chembl/__init__.py

@@ -1,9 +1,15 @@
 """Resources from ChEMBL."""
 
+from .chembl_cell import ChEMBLCellGetter
 from .chembl_compound import ChEMBLCompoundGetter
+from .chembl_mechanism import ChEMBLMechanismGetter
 from .chembl_target import ChEMBLTargetGetter
+from .chembl_tissue import ChEMBLTissueGetter
 
 __all__ = [
+    "ChEMBLCellGetter",
     "ChEMBLCompoundGetter",
+    "ChEMBLMechanismGetter",
     "ChEMBLTargetGetter",
+    "ChEMBLTissueGetter",
 ]

pyobo/sources/chembl/chembl_cell.py

@@ -0,0 +1,94 @@
+"""Converter for ChEMBL cells."""
+
+import logging
+from collections.abc import Iterable
+
+import chembl_downloader
+
+from pyobo.struct import Obo, Reference, Term
+from pyobo.struct.typedef import derives_from_organism, exact_match
+
+__all__ = [
+    "ChEMBLCellGetter",
+]
+
+logger = logging.getLogger(__name__)
+
+PREFIX = "chembl.cell"
+
+
+class ChEMBLCellGetter(Obo):
+    """An ontology representation of ChEMBL cells."""
+
+    ontology = PREFIX
+    bioversions_key = "chembl"
+    typedefs = [exact_match, derives_from_organism]
+
+    def iter_terms(self, force: bool = False) -> Iterable[Term]:
+        """Iterate over terms in the ontology."""
+        return iter_terms(version=self._version_or_raise)
+
+
+QUERY = """\
+SELECT
+    CHEMBL_ID,
+    CELL_NAME,
+    CELL_DESCRIPTION,
+    CELL_SOURCE_TISSUE,
+    CELL_SOURCE_TAX_ID,
+    CLO_ID,
+    EFO_ID,
+    CELLOSAURUS_ID,
+    CL_LINCS_ID,
+    CELL_ONTOLOGY_ID
+FROM CELL_DICTIONARY
+"""
+
+
+def iter_terms(version: str | None = None) -> Iterable[Term]:
+    """Iterate over ChEMBL cell terms."""
+    with chembl_downloader.cursor(version=version) as cursor:
+        cursor.execute(QUERY)
+        for (
+            chembl_id,
+            name,
+            desc,
+            _source_tissue,
+            taxid,
+            clo,
+            efo,
+            cellosaurus,
+            lincs,
+            cl,
+        ) in cursor.fetchall():
+            term = Term(
+                reference=Reference(prefix=PREFIX, identifier=chembl_id, name=name),
+                definition=desc if desc and desc != name else None,
+            )
+            if taxid:
+                term.append_relationship(
+                    derives_from_organism, Reference(prefix="ncbitaxon", identifier=taxid)
+                )
+            # TODO how to annotate tissue, via TISSUE_DICTIONARY
+            if clo:
+                term.append_exact_match(
+                    Reference(prefix="clo", identifier=clo.removeprefix("CLO_"))
+                )
+            if efo:
+                term.append_exact_match(
+                    Reference(prefix="efo", identifier=efo.removeprefix("EFO_").removeprefix("EFO"))
+                )
+            if cellosaurus:
+                term.append_exact_match(
+                    Reference(prefix="cellosaurus", identifier=cellosaurus.removeprefix("CVCL_"))
+                )
+            if lincs:
+                # with LCL- included!
+                term.append_exact_match(Reference(prefix="lincs.cell", identifier=lincs))
+            if cl:
+                term.append_exact_match(Reference(prefix="cl", identifier=cl.removeprefix("CL_")))
+            yield term
+
+
+if __name__ == "__main__":
+    ChEMBLCellGetter.cli()

pyobo/sources/chembl/chembl_mechanism.py

@@ -0,0 +1,81 @@
+"""Converter for ChEMBL mechanisms."""
+
+import logging
+from collections.abc import Iterable
+
+import chembl_downloader
+
+from pyobo.struct import CHARLIE_TERM, PYOBO_INJECTED, Obo, Term
+from pyobo.struct.typedef import exact_match
+
+__all__ = [
+    "ChEMBLMechanismGetter",
+]
+
+logger = logging.getLogger(__name__)
+
+PREFIX = "chembl.mechanism"
+QUERY = "SELECT * from ACTION_TYPE"
+
+ROOT = (
+    Term.default(PREFIX, "mechanism", name="mechanism")
+    .append_contributor(CHARLIE_TERM)
+    .append_comment(PYOBO_INJECTED)
+)
+
+
+class ChEMBLMechanismGetter(Obo):
+    """An ontology representation of ChEMBL mechanisms."""
+
+    ontology = PREFIX
+    bioversions_key = "chembl"
+    typedefs = [exact_match]
+    root_terms = [ROOT.reference]
+
+    def iter_terms(self, force: bool = False) -> Iterable[Term]:
+        """Iterate over terms in the ontology."""
+        return iter_terms(version=self._version_or_raise)
+
+
+def normalize_chembl_mechanism(name: str) -> str:
+    """Normalize a ChEMBL mechanism name into an identifier."""
+    return name.lower().replace(" ", "-")
+
+
+def _norm_name(name: str) -> str:
+    return name.lower().replace("rnai ", "RNAi ")
+
+
+def get_pattern(version: str | None = None) -> str:
+    """Get a pattern."""
+    df = chembl_downloader.query("SELECT action_type from ACTION_TYPE", version=version)
+    parts = "|".join(sorted(normalize_chembl_mechanism(name) for (name,) in df.values))
+    return f"^[{parts}]$"
+
+
+def iter_terms(version: str) -> Iterable[Term]:
+    """Iterate over ChEMBL mechanisms."""
+    df = chembl_downloader.query(QUERY, version=version)
+    terms = {}
+    parents = {}
+    for name, _description, parent in df.values:
+        identifier = normalize_chembl_mechanism(name)
+        terms[name] = Term.from_triple(prefix=PREFIX, identifier=identifier, name=_norm_name(name))
+        if name != parent:  # protect against "other" which is a child of itself
+            parents[name] = parent
+    for child, parent in parents.items():
+        terms[child].append_parent(terms[parent])
+
+    # these are the three top-level things in the hierarchy, which
+    # we annotate onto a dummy parent term
+    for name in [
+        "POSITIVE MODULATOR",
+        "NEGATIVE MODULATOR",
+        "OTHER",
+    ]:
+        terms[name].append_parent(ROOT)
+    yield from terms.values()
+
+
+if __name__ == "__main__":
+    ChEMBLMechanismGetter.cli()