feat!: update to vrs 2.0 models

pyproject.toml

@@ -26,10 +26,9 @@ description = "Computable object representation and validation for gene fusions"
 license = {file = "LICENSE"}
 dependencies = [
     "pydantic == 2.*",
-    "ga4gh.vrsatile.pydantic ~=0.2.0",
-    "ga4gh.vrs ~=0.8.1",
+    "ga4gh.vrs ~=2.0.0a8",
     "biocommons.seqrepo",
-    "gene-normalizer ~=0.1.40-dev1",
+    "gene-normalizer ~=0.4.0",
     "cool-seq-tool ~=0.5.0",
 ]
 dynamic=["version"]

src/fusor/fusor.py

@@ -8,18 +8,13 @@
 from cool_seq_tool.app import CoolSeqTool
 from cool_seq_tool.schemas import ResidueMode
 from ga4gh.core import ga4gh_identify
+from ga4gh.core.domain_models import Gene
 from ga4gh.vrs import models
-from ga4gh.vrsatile.pydantic.vrs_models import (
-    CURIE,
-    Number,
-    SequenceInterval,
-    SequenceLocation,
-    VRSTypes,
-)
-from ga4gh.vrsatile.pydantic.vrsatile_models import GeneDescriptor, LocationDescriptor
+from ga4gh.vrs.models import SequenceLocation
 from gene.database import AbstractDatabase as GeneDatabase
 from gene.database import create_db
 from gene.query import QueryHandler
+from gene.schemas import CURIE
 from pydantic import ValidationError
 
 from fusor.exceptions import FUSORParametersException, IDTranslationException
@@ -100,7 +95,7 @@ def _contains_element_type(kwargs: dict, elm_type: StructuralElementType) -> boo
     def fusion(self, fusion_type: FusionType | None = None, **kwargs) -> Fusion:
         """Construct fusion object.
 
-        :param fusion_type: explicitly specify fusion type. Unecessary if providing
+        :param fusion_type: explicitly specify fusion type. Unnecessary if providing
             fusion object in keyword args that includes ``type`` attribute.
         :return: constructed fusion object if successful
         :raise: FUSORParametersException if fusion type unable to be determined,
@@ -292,7 +287,7 @@ async def transcript_segment_element(
                 exon_end=genomic_data.exon_end,
                 exon_end_offset=genomic_data.exon_end_offset,
                 gene_descriptor=normalized_gene_response[0],
-                element_genomic_start=self._location_descriptor(
+                element_genomic_start=self._sequence_location(
                     genomic_data.start,
                     genomic_data.start + 1,
                     genomic_data.chr,
@@ -301,7 +296,7 @@ async def transcript_segment_element(
                 )
                 if genomic_data.start
                 else None,
-                element_genomic_end=self._location_descriptor(
+                element_genomic_end=self._sequence_location(
                     genomic_data.end,
                     genomic_data.end + 1,
                     genomic_data.chr,
@@ -361,7 +356,7 @@ def templated_sequence_element(
         if residue_mode == ResidueMode.RESIDUE:
             start -= 1
 
-        region = self._location_descriptor(
+        region = self._sequence_location(
             start,
             end,
             sequence_id,
@@ -468,7 +463,7 @@ def functional_domain(
         if not gene_descr:
             return None, warning
 
-        loc_descr = self._location_descriptor(
+        loc_descr = self._sequence_location(
             start, end, sequence_id, seq_id_target_namespace=seq_id_target_namespace
         )
 
@@ -519,15 +514,15 @@ def regulatory_element(
             _logger.warning(msg)
             return None, msg
 
-    def _location_descriptor(
+    def _sequence_location(
         self,
         start: int,
         end: int,
         sequence_id: str,
         label: str | None = None,
         seq_id_target_namespace: str | None = None,
         use_location_id: bool = False,
-    ) -> LocationDescriptor:
+    ) -> SequenceLocation:
         """Create location descriptor
 
         :param start: Start position
@@ -564,22 +559,11 @@ def _location_descriptor(
             else:
                 sequence_id = seq_id
 
-        location = SequenceLocation(
+        return SequenceLocation(
             sequence_id=sequence_id,
-            interval=SequenceInterval(start=Number(value=start), end=Number(value=end)),
+            start=start, end=end,
         )
 
-        if use_location_id:
-            _id = self._location_id(location.model_dump())
-        else:
-            quote_id = quote(label) if label else quote(seq_id_input)
-            _id = f"fusor.location_descriptor:{quote_id}"
-
-        location_descr = LocationDescriptor(id=_id, location=location)
-
-        if label:
-            location_descr.label = label
-        return location_descr
 
     def add_additional_fields(
         self,
@@ -636,7 +620,7 @@ def add_location_id(self, fusion: Fusion) -> Fusion:
                 ]:
                     if element_genomic:
                         location = element_genomic.location
-                        if location.type == VRSTypes.SEQUENCE_LOCATION.value:
+                        if location.type == SequenceLocation:
                             location_id = self._location_id(location.model_dump())
                             element_genomic.location_id = location_id
         if isinstance(fusion, CategoricalFusion) and fusion.critical_functional_domains:
@@ -648,7 +632,7 @@ def add_location_id(self, fusion: Fusion) -> Fusion:
             element = fusion.regulatory_element
             if element.feature_location:
                 location = element.feature_location
-                if location.type == VRSTypes.SEQUENCE_LOCATION.value:
+                if location.type == SequenceLocation:
                     location_id = self._location_id(location.model_dump())
                     element.feature_location.location_id = location_id
         return fusion
@@ -674,7 +658,7 @@ def add_translated_sequence_id(
         for element in fusion.structural_elements:
             if isinstance(element, TemplatedSequenceElement):
                 location = element.region.location
-                if location.type == VRSTypes.SEQUENCE_LOCATION.value:
+                if location.type == SequenceLocation:
                     try:
                         new_id = translate_identifier(
                             self.seqrepo, location.sequence_id, target_namespace
@@ -690,7 +674,7 @@ def add_translated_sequence_id(
                 ]:
                     if loc_descr:
                         location = loc_descr.location
-                        if location.type == VRSTypes.SEQUENCE_LOCATION.value:
+                        if location.type == SequenceLocation:
                             try:
                                 new_id = translate_identifier(
                                     self.seqrepo, location.sequence_id, target_namespace
@@ -716,6 +700,7 @@ def add_translated_sequence_id(
                     domain.sequence_location.location.sequence_id = new_id
         return fusion
 
+    # TODO: should this be adding to the gene extensions or something instead?
     def add_gene_descriptor(self, fusion: Fusion) -> Fusion:
         """Add additional fields to ``gene_descriptor`` in fusion object
 
@@ -730,24 +715,24 @@ def add_gene_descriptor(self, fusion: Fusion) -> Fusion:
             for obj in prop:
                 if "gene_descriptor" in obj.model_fields:
                     label = obj.gene_descriptor.label
-                    norm_gene_descr, _ = self._normalized_gene_descriptor(
+                    norm_gene_descr, _ = self._normalized_gene(
                         label, use_minimal_gene_descr=False
                     )
                     if norm_gene_descr:
                         obj.gene_descriptor = norm_gene_descr
         if fusion.regulatory_element and fusion.regulatory_element.associated_gene:
             reg_el = fusion.regulatory_element
             label = reg_el.associated_gene.label
-            norm_gene_descr, _ = self._normalized_gene_descriptor(
+            norm_gene_descr, _ = self._normalized_gene(
                 label, use_minimal_gene_descr=False
             )
             if norm_gene_descr:
                 reg_el.associated_gene = norm_gene_descr
         return fusion
 
-    def _normalized_gene_descriptor(
+    def _normalized_gene(
         self, query: str, use_minimal_gene_descr: bool = True
-    ) -> tuple[GeneDescriptor | None, str | None]:
+    ) -> tuple[Gene | None, str | None]:
         """Return gene descriptor from normalized response.
 
         :param query: Gene query
@@ -761,7 +746,8 @@ def _normalized_gene_descriptor(
         if gene_norm_resp.match_type:
             gene_descr = gene_norm_resp.gene_descriptor
             if use_minimal_gene_descr:
-                gene_descr = GeneDescriptor(
+                # TODO: how to handle gene_id here? add to extensions??
+                gene_descr = Gene(
                     id=gene_descr.id, gene_id=gene_descr.gene_id, label=gene_descr.label
                 )
             return gene_descr, None

src/fusor/models.py

@@ -4,13 +4,9 @@
 from enum import Enum
 from typing import Any, Literal
 
-from ga4gh.vrsatile.pydantic import return_value
-from ga4gh.vrsatile.pydantic.vrsatile_models import (
-    CURIE,
-    GeneDescriptor,
-    LocationDescriptor,
-    SequenceDescriptor,
-)
+from ga4gh.core.domain_models import Gene
+from ga4gh.vrs.models import SequenceLocation
+from gene.schemas import CURIE
 from pydantic import (
     BaseModel,
     ConfigDict,
@@ -34,7 +30,8 @@ class FUSORTypes(str, Enum):
     TRANSCRIPT_SEGMENT_ELEMENT = "TranscriptSegmentElement"
     TEMPLATED_SEQUENCE_ELEMENT = "TemplatedSequenceElement"
     LINKER_SEQUENCE_ELEMENT = "LinkerSequenceElement"
-    GENE_ELEMENT = "GeneElement"
+    # TODO: I'm not sure if this needs to still be here or not
+    GENE = "Gene"
     UNKNOWN_GENE_ELEMENT = "UnknownGeneElement"
     MULTIPLE_POSSIBLE_GENES_ELEMENT = "MultiplePossibleGenesElement"
     REGULATORY_ELEMENT = "RegulatoryElement"
@@ -63,11 +60,12 @@ class FunctionalDomain(BaseModel):
 
     type: Literal[FUSORTypes.FUNCTIONAL_DOMAIN] = FUSORTypes.FUNCTIONAL_DOMAIN
     status: DomainStatus
-    associated_gene: GeneDescriptor
+    associated_gene: Gene
     id: CURIE | None = Field(None, alias="_id")
     label: StrictStr | None = None
-    sequence_location: LocationDescriptor | None = None
+    sequence_location: SequenceLocation | None = None
 
+    # TODO: is this obsolete now that vrsatile has been removed?
     _get_id_val = field_validator("id")(return_value)
 
     model_config = ConfigDict(
@@ -107,7 +105,7 @@ class StructuralElementType(str, Enum):
     TRANSCRIPT_SEGMENT_ELEMENT = FUSORTypes.TRANSCRIPT_SEGMENT_ELEMENT.value
     TEMPLATED_SEQUENCE_ELEMENT = FUSORTypes.TEMPLATED_SEQUENCE_ELEMENT.value
     LINKER_SEQUENCE_ELEMENT = FUSORTypes.LINKER_SEQUENCE_ELEMENT.value
-    GENE_ELEMENT = FUSORTypes.GENE_ELEMENT.value
+    GENE_ELEMENT = Gene
     UNKNOWN_GENE_ELEMENT = FUSORTypes.UNKNOWN_GENE_ELEMENT.value
     MULTIPLE_POSSIBLE_GENES_ELEMENT = FUSORTypes.MULTIPLE_POSSIBLE_GENES_ELEMENT.value
 
@@ -129,9 +127,9 @@ class TranscriptSegmentElement(BaseStructuralElement):
     exon_start_offset: StrictInt | None = 0
     exon_end: StrictInt | None = None
     exon_end_offset: StrictInt | None = 0
-    gene_descriptor: GeneDescriptor
-    element_genomic_start: LocationDescriptor | None = None
-    element_genomic_end: LocationDescriptor | None = None
+    gene: Gene
+    element_genomic_start: SequenceLocation | None = None
+    element_genomic_end: SequenceLocation | None = None
 
     @model_validator(mode="before")
     def check_exons(cls, values):
@@ -216,7 +214,7 @@ class LinkerElement(BaseStructuralElement, extra="forbid"):
     type: Literal[FUSORTypes.LINKER_SEQUENCE_ELEMENT] = (
         FUSORTypes.LINKER_SEQUENCE_ELEMENT
     )
-    linker_sequence: SequenceDescriptor
+    linker_sequence: SequenceLocation
 
     @field_validator("linker_sequence", mode="before")
     def validate_sequence(cls, v):
@@ -226,7 +224,7 @@ def validate_sequence(cls, v):
                 v["sequence"] = v["sequence"].upper()
             except KeyError as e:
                 raise TypeError from e
-        elif isinstance(v, SequenceDescriptor):
+        elif isinstance(v, SequenceLocation):
             v.sequence = v.sequence.upper()
         else:
             raise TypeError
@@ -264,7 +262,7 @@ class TemplatedSequenceElement(BaseStructuralElement):
     type: Literal[FUSORTypes.TEMPLATED_SEQUENCE_ELEMENT] = (
         FUSORTypes.TEMPLATED_SEQUENCE_ELEMENT
     )
-    region: LocationDescriptor
+    region: SequenceLocation
     strand: Strand
 
     model_config = ConfigDict(
@@ -292,27 +290,6 @@ class TemplatedSequenceElement(BaseStructuralElement):
     )
 
 
-class GeneElement(BaseStructuralElement):
-    """Define Gene Element class."""
-
-    type: Literal[FUSORTypes.GENE_ELEMENT] = FUSORTypes.GENE_ELEMENT
-    gene_descriptor: GeneDescriptor
-
-    model_config = ConfigDict(
-        json_schema_extra={
-            "example": {
-                "type": "GeneElement",
-                "gene_descriptor": {
-                    "id": "gene:BRAF",
-                    "gene_id": "hgnc:1097",
-                    "label": "BRAF",
-                    "type": "GeneDescriptor",
-                },
-            }
-        },
-    )
-
-
 class UnknownGeneElement(BaseStructuralElement):
     """Define UnknownGene class. This is primarily intended to represent a
     partner in the result of a fusion partner-agnostic assay, which identifies
@@ -388,8 +365,8 @@ class RegulatoryElement(BaseModel):
     type: Literal[FUSORTypes.REGULATORY_ELEMENT] = FUSORTypes.REGULATORY_ELEMENT
     regulatory_class: RegulatoryClass
     feature_id: str | None = None
-    associated_gene: GeneDescriptor | None = None
-    feature_location: LocationDescriptor | None = None
+    associated_gene: Gene | None = None
+    feature_location: SequenceLocation | None = None
 
     _get_ref_id_val = field_validator("feature_id")(return_value)
 
@@ -614,7 +591,7 @@ class Assay(BaseModelForbidExtra):
 
 AssayedFusionElements = list[
     TranscriptSegmentElement
-    | GeneElement
+    | Gene
     | TemplatedSequenceElement
     | LinkerElement
     | UnknownGeneElement
@@ -682,13 +659,10 @@ class AssayedFusion(AbstractFusion):
                 },
                 "structural_elements": [
                     {
-                        "type": "GeneElement",
-                        "gene_descriptor": {
-                            "id": "gene:EWSR1",
-                            "gene_id": "hgnc:3058",
-                            "label": "EWSR1",
-                            "type": "GeneDescriptor",
-                        },
+                        "type": "Gene",
+                        "id": "gene:EWSR1",
+                        "gene_id": "hgnc:3058",
+                        "label": "EWSR1",
                     },
                     {"type": "UnknownGeneElement"},
                 ],
@@ -699,7 +673,7 @@ class AssayedFusion(AbstractFusion):
 
 CategoricalFusionElements = list[
     TranscriptSegmentElement
-    | GeneElement
+    | Gene
     | TemplatedSequenceElement
     | LinkerElement
     | MultiplePossibleGenesElement
@@ -781,13 +755,10 @@ class CategoricalFusion(AbstractFusion):
                         },
                     },
                     {
-                        "type": "GeneElement",
-                        "gene_descriptor": {
-                            "id": "gene:ALK",
-                            "type": "GeneDescriptor",
-                            "gene_id": "hgnc:427",
-                            "label": "ALK",
-                        },
+                        "type": "Gene",
+                        "id": "gene:ALK",
+                        "gene_id": "hgnc:427",
+                        "label": "ALK",
                     },
                 ],
                 "regulatory_element": {

src/fusor/nomenclature.py

@@ -1,7 +1,6 @@
 """Provide helper methods for fusion nomenclature generation."""
 
 from biocommons.seqrepo.seqrepo import SeqRepo
-from ga4gh.vrsatile.pydantic.vrs_models import SequenceLocation
 
 from fusor.exceptions import IDTranslationException
 from fusor.models import (