Merge pull request #30 from monarch-initiative/29-remove-dependency-o…

…n-ontobio

29 remove dependency on ontobio

pyproject.toml

@@ -13,8 +13,7 @@ phenopackets = "^2.0.2"
 polars = "^0.19"
 oaklib = "^0.5.1"
 click = "^8.1.3"
-ontobio = "^2.8.5"
-pheval = {git = "https://github.com/monarch-initiative/pheval.git"}
+pheval = "^0.3.6"
 
 [tool.poetry.dev-dependencies]
 pytest = "^7.2.0"

src/phenotype2phenopacket/create/create.py

@@ -2,7 +2,6 @@
 
 import polars as pl
 from oaklib.implementations import ProntoImplementation
-from ontobio import Ontology
 
 from phenotype2phenopacket.utils.phenopacket_utils import (
     PhenotypeAnnotationToPhenopacketConverter,
@@ -12,15 +11,13 @@
 from phenotype2phenopacket.utils.utils import (
     filter_diseases,
     load_ontology,
-    load_ontology_factory,
     return_phenotype_annotation_data,
 )
 
 
 def create_synthetic_patient_phenopacket(
     human_phenotype_ontology: ProntoImplementation,
     omim_disease: pl.DataFrame,
-    ontology_factory: Ontology,
     output_dir: Path,
     pt_id: str,
     hpoa_version: str,
@@ -31,14 +28,12 @@ def create_synthetic_patient_phenopacket(
     Args:
         human_phenotype_ontology: An instance of ProntoImplementation containing the loaded HPO.
         omim_disease (pl.DataFrame): DataFrame containing phenotype entries for a specific OMIM disease.
-        ontology_factory: Created HPO ontology from OntologyFactory
         output_dir (Path): The directory path to write the generated phenopacket.
+        pt_id (str): The patient ID.
         hpoa_version (str): The version of the Human Phenotype Ontology Annotation.
 
     """
-    synthetic_patient_generator = SyntheticPatientGenerator(
-        omim_disease, human_phenotype_ontology, ontology_factory
-    )
+    synthetic_patient_generator = SyntheticPatientGenerator(omim_disease, human_phenotype_ontology)
     patient_terms = synthetic_patient_generator.patient_term_annotation_set()
     phenopacket_file = PhenotypeAnnotationToPhenopacketConverter(
         human_phenotype_ontology
@@ -70,15 +65,14 @@ def create_synthetic_patients(
     """
     phenotype_annotation_data = return_phenotype_annotation_data(phenotype_annotation)
     human_phenotype_ontology = load_ontology()
-    ontology_factory = load_ontology_factory()
     grouped_omim_diseases = filter_diseases(
         num_disease, omim_id, omim_id_list, phenotype_annotation_data
     )
     for omim_disease in grouped_omim_diseases:
         create_synthetic_patient_phenopacket(
             human_phenotype_ontology,
             omim_disease,
-            ontology_factory,
             output_dir,
+            None,
             phenotype_annotation_data.version,
         )

src/phenotype2phenopacket/utils/phenopacket_utils.py

@@ -11,7 +11,6 @@
 import polars as pl
 from google.protobuf.timestamp_pb2 import Timestamp
 from oaklib.implementations import ProntoImplementation
-from ontobio import Ontology
 from phenopackets import (
     Age,
     Diagnosis,
@@ -150,21 +149,17 @@ def write_phenopacket(phenopacket: Phenopacket, output_file: Path) -> None:
 class SyntheticPatientGenerator:
     """Class for generating synthetic patients."""
 
-    def __init__(
-        self, disease_df: pl.DataFrame, ontology: ProntoImplementation, ontology_factory: Ontology
-    ):
+    def __init__(self, disease_df: pl.DataFrame, ontology: ProntoImplementation):
         """
         Initialise the SyntheticPatientGenerator class
 
         Args:
             disease_df (pl.DataFrame): The dataframe containing the annotation data for a specific disease.
             ontology (ProntoImplementation): An instance of ProntoImplementation containing the loaded HPO.
-            ontology_factory (Ontology): Created ontology from OntologyFactory
 
         """
         self.disease_df = disease_df
         self.ontology = ontology
-        self.ontology_factory = ontology_factory
         self.lower_age = 0
         self.upper_age = 0
         self.filtered_df = []
@@ -452,9 +447,9 @@ def get_children_of_term(self, phenotype_entry: dict, steps: int) -> dict:
         """
         term_id = phenotype_entry["hpo_id"]
         for _i in range(steps):
-            descendants = self.ontology_factory.children(term_id)
+            descendants = list(self.ontology.incoming_relationships(term_id))
             if descendants:
-                term_id = self.secret_rand.choice(descendants)
+                term_id = self.secret_rand.choice(descendants)[1]
             else:
                 break
         phenotype_entry["hpo_id"] = term_id
@@ -602,6 +597,7 @@ def create_individual(
         Args:
             onset_range (OnsetTerm, optional): An OnsetTerm object representing the age range of onset.
             Defaults to None.
+            pt_id (str, optional): An identifier for the patient.
 
         Returns:
             Individual: An instance of the Individual class.

src/phenotype2phenopacket/utils/utils.py

@@ -5,9 +5,7 @@
 
 import pandas as pd
 import polars as pl
-from oaklib import OntologyResource
-from oaklib.implementations import ProntoImplementation
-from ontobio import Ontology, OntologyFactory
+from oaklib import get_adapter
 
 
 def is_float(element: any) -> bool:
@@ -65,21 +63,9 @@ def load_ontology():
     Load the Human Phenotype Ontology (HPO).
 
     Returns:
-        ProntoImplementation: An instance of ProntoImplementation containing the loaded HPO.
+        An instantiated interface.
     """
-    resource = OntologyResource(slug="hp.obo", local=False)
-    return ProntoImplementation(resource)
-
-
-def load_ontology_factory() -> Ontology:
-    """
-    Load human phenotype ontology factory.
-
-    Returns:
-        Ontology: An instance of the hp Ontology class.
-    """
-    ontology_factory = OntologyFactory()
-    return ontology_factory.create("hp")
+    return get_adapter("sqlite:obo:hp")
 
 
 def read_hgnc_data(hgnc_data_file: Path) -> pd.DataFrame:
@@ -114,7 +100,7 @@ def read_phenotype_annotation_file(phenotype_annotation_file_path: Path) -> pl.D
 
 def filter_phenotype_annotation(phenotype_annotation_file_path: Path) -> pl.DataFrame:
     """
-    Filter the phenotype annotation, retaining data for OMIM diseases describing phenotypye information.
+    Filter the phenotype annotation, retaining data for OMIM diseases describing phenotype information.
     Args:
         phenotype_annotation_file_path (Path): The path to the phenotype annotation file.
 

tests/test_phenopacket_utils.py

@@ -34,7 +34,7 @@
     SyntheticPatientGenerator,
     create_phenopacket_file_name_from_disease,
 )
-from phenotype2phenopacket.utils.utils import load_ontology, load_ontology_factory
+from phenotype2phenopacket.utils.utils import load_ontology
 
 disease_df = pl.from_dicts(
     [
@@ -298,7 +298,6 @@ def setUpClass(cls) -> None:
         cls.synthetic_patient_generator = SyntheticPatientGenerator(
             disease_df=disease_df,
             ontology=load_ontology(),
-            ontology_factory=load_ontology_factory(),
         )
 
     def tearDown(self) -> None:
@@ -746,7 +745,7 @@ def mock_choice(lst):
                     "database_id": "OMIM:612567",
                     "disease_name": "Inflammatory bowel disease 25, early onset, autosomal recessive",
                     "qualifier": None,
-                    "hpo_id": "HP:0009789",
+                    "hpo_id": "HP:0001047",
                     "reference": "PMID:21519361",
                     "evidence": "PCS",
                     "onset": None,