Merge pull request #30 from VariantEffect/mavedb-dev

MaveDB Mapping v2024.1.2

Author: bencap

src/api/routers/map.py

@@ -1,4 +1,6 @@
 """"Provide mapping router"""
+from pathlib import Path
+
 from cool_seq_tool.schemas import AnnotationLayer
 from fastapi import APIRouter, HTTPException
 from fastapi.responses import JSONResponse
@@ -17,6 +19,7 @@
     get_raw_scoreset_metadata,
     get_scoreset_metadata,
     get_scoreset_records,
+    with_mavedb_score_set,
 )
 from dcd_mapping.resource_utils import ResourceAcquisitionError
 from dcd_mapping.schemas import ScoreAnnotation, ScoresetMapping, VrsVersion
@@ -29,7 +32,8 @@
 
 
 @router.post(path="/map/{urn}", status_code=200, response_model=ScoresetMapping)
-async def map_scoreset(urn: str) -> ScoresetMapping:
+@with_mavedb_score_set
+async def map_scoreset(urn: str, store_path: Path | None = None) -> ScoresetMapping:
     """Perform end-to-end mapping for a scoreset.
 
     :param urn: identifier for a scoreset.
@@ -38,8 +42,8 @@ async def map_scoreset(urn: str) -> ScoresetMapping:
     :param silent: if True, suppress console information output
     """
     try:
-        metadata = get_scoreset_metadata(urn)
-        records = get_scoreset_records(urn, True)
+        metadata = get_scoreset_metadata(urn, store_path)
+        records = get_scoreset_records(urn, True, store_path)
     except ScoresetNotSupportedError as e:
         return ScoresetMapping(
             metadata=None,
@@ -49,6 +53,14 @@ async def map_scoreset(urn: str) -> ScoresetMapping:
         msg = f"Unable to acquire resource from MaveDB: {e}"
         raise HTTPException(status_code=500, detail=msg) from e
 
+    if not records:
+        return JSONResponse(
+            content=ScoresetMapping(
+                metadata=metadata,
+                error_message="Score set contains no variants to map",
+            ).model_dump(exclude_none=True)
+        )
+
     try:
         alignment_result = align(metadata, True)
     except BlatNotFoundError as e:
@@ -108,7 +120,7 @@ async def map_scoreset(urn: str) -> ScoresetMapping:
         )
 
     try:
-        raw_metadata = get_raw_scoreset_metadata(urn)
+        raw_metadata = get_raw_scoreset_metadata(urn, store_path)
         preferred_layers = {
             _set_scoreset_layer(urn, vrs_results),
         }
@@ -124,7 +136,7 @@ async def map_scoreset(urn: str) -> ScoresetMapping:
         for layer in preferred_layers:
             reference_sequences[layer][
                 "computed_reference_sequence"
-            ] = _get_computed_reference_sequence(urn, layer, transcript)
+            ] = _get_computed_reference_sequence(metadata, layer, transcript)
             reference_sequences[layer][
                 "mapped_reference_sequence"
             ] = _get_mapped_reference_sequence(layer, transcript, alignment_result)

src/dcd_mapping/annotate.py

@@ -29,7 +29,6 @@
     get_seqrepo,
     get_vrs_id_from_identifier,
 )
-from dcd_mapping.mavedb_data import get_raw_scoreset_metadata, get_scoreset_metadata
 from dcd_mapping.resource_utils import LOCAL_STORE_PATH
 from dcd_mapping.schemas import (
     AlignmentResult,
@@ -409,7 +408,7 @@ def annotate(
 
 
 def _get_computed_reference_sequence(
-    ss: str,
+    metadata: ScoresetMetadata,
     layer: AnnotationLayer,
     tx_output: TxSelectResult | None = None,
 ) -> ComputedReferenceSequence:
@@ -429,7 +428,6 @@ def _get_computed_reference_sequence(
             sequence_type=TargetSequenceType.PROTEIN,
             sequence_id=seq_id,
         )
-    metadata = get_scoreset_metadata(ss)
     seq_id = f"ga4gh:SQ.{sha512t24u(metadata.target_sequence.encode('ascii'))}"
     return ComputedReferenceSequence(
         sequence=metadata.target_sequence,
@@ -516,7 +514,7 @@ def write_scoreset_mapping_to_json(
 
 
 def save_mapped_output_json(
-    urn: str,
+    metadata: ScoresetMetadata,
     mappings: list[ScoreAnnotationWithLayer],
     align_result: AlignmentResult,
     tx_output: TxSelectResult | None,
@@ -533,10 +531,9 @@ def save_mapped_output_json(
         <dcd_mapping_data_dir>/urn:mavedb:00000XXX-X-X_mapping_<ISO8601 datetime>.json
     :return: output location
     """
-    metadata = get_raw_scoreset_metadata(urn)
     if preferred_layer_only:
         preferred_layers = {
-            _set_scoreset_layer(urn, mappings),
+            _set_scoreset_layer(metadata.urn, mappings),
         }
     else:
         preferred_layers = {mapping.annotation_layer for mapping in mappings}
@@ -549,20 +546,10 @@ def save_mapped_output_json(
     for layer in preferred_layers:
         reference_sequences[layer][
             "computed_reference_sequence"
-        ] = _get_computed_reference_sequence(urn, layer, tx_output)
+        ] = _get_computed_reference_sequence(metadata, layer, tx_output)
         reference_sequences[layer][
             "mapped_reference_sequence"
         ] = _get_mapped_reference_sequence(layer, tx_output, align_result)
-    # except Exception as e:
-    #     _logger.warning(
-    #         str(e)
-    #     )
-    #     output = ScoresetMapping(
-    #         metadata=metadata,
-    #         error_message = str(e).strip("'")
-    #     )
-
-    #     return write_scoreset_mapping_to_json
 
     mapped_scores: list[ScoreAnnotation] = []
     for m in mappings:
@@ -573,7 +560,7 @@ def save_mapped_output_json(
             mapped_scores.append(ScoreAnnotation(**m.model_dump()))
 
     output = ScoresetMapping(
-        metadata=metadata,
+        metadata=metadata.model_dump(),
         computed_protein_reference_sequence=reference_sequences[
             AnnotationLayer.PROTEIN
         ]["computed_reference_sequence"],
@@ -589,4 +576,4 @@ def save_mapped_output_json(
         mapped_scores=mapped_scores,
     )
 
-    return write_scoreset_mapping_to_json(urn, output, output_path)
+    return write_scoreset_mapping_to_json(metadata.urn, output, output_path)

src/dcd_mapping/main.py

@@ -24,6 +24,7 @@
     ScoresetNotSupportedError,
     get_scoreset_metadata,
     get_scoreset_records,
+    with_mavedb_score_set,
 )
 from dcd_mapping.resource_utils import ResourceAcquisitionError
 from dcd_mapping.schemas import (
@@ -264,7 +265,7 @@ async def map_scoreset(
         return
     try:
         final_output = save_mapped_output_json(
-            metadata.urn,
+            metadata,
             vrs_results,
             alignment_result,
             transcript,
@@ -287,12 +288,14 @@ async def map_scoreset(
     _emit_info(f"Annotated scores saved to: {final_output}.", silent)
 
 
+@with_mavedb_score_set
 async def map_scoreset_urn(
     urn: str,
     output_path: Path | None = None,
     vrs_version: VrsVersion = VrsVersion.V_2,
     prefer_genomic: bool = False,
     silent: bool = True,
+    store_path: Path | None = None,
 ) -> None:
     """Perform end-to-end mapping for a scoreset.
 
@@ -302,8 +305,8 @@ async def map_scoreset_urn(
     :param silent: if True, suppress console information output
     """
     try:
-        metadata = get_scoreset_metadata(urn)
-        records = get_scoreset_records(urn, silent)
+        metadata = get_scoreset_metadata(urn, store_path)
+        records = get_scoreset_records(urn, silent, store_path)
     except ScoresetNotSupportedError as e:
         _emit_info(f"Score set not supported: {e}", silent, logging.ERROR)
         final_output = write_scoreset_mapping_to_json(
@@ -321,6 +324,20 @@ async def map_scoreset_urn(
         _logger.critical(msg)
         click.echo(f"Error: {msg}")
         raise e
+
+    if not records:
+        _emit_info("Score set contains no variants to map", silent, logging.ERROR)
+        final_output = write_scoreset_mapping_to_json(
+            urn,
+            ScoresetMapping(
+                metadata=metadata,
+                error_message="Score set contains no variants to map",
+            ),
+            output_path,
+        )
+        _emit_info(f"Score set mapping output saved to: {final_output}.", silent)
+        return
+
     await map_scoreset(
         metadata, records, output_path, vrs_version, prefer_genomic, silent
     )

src/dcd_mapping/mavedb_data.py

@@ -2,11 +2,14 @@
 
 Much of this can/should be replaced by the ``mavetools`` library? (and/or ``wags-tails``.)
 """
+
 import csv
 import json
 import logging
 import tempfile
 import zipfile
+from collections.abc import Callable
+from functools import wraps
 from pathlib import Path
 from typing import Any
 
@@ -20,7 +23,7 @@
     authentication_header,
     http_download,
 )
-from dcd_mapping.schemas import ScoreRow, ScoresetMetadata, UniProtRef
+from dcd_mapping.schemas import ScoreRow, ScoresetMapping, ScoresetMetadata, UniProtRef
 
 __all__ = [
     "get_scoreset_urns",
@@ -135,6 +138,7 @@ def get_raw_scoreset_metadata(
     """
     if not dcd_mapping_dir:
         dcd_mapping_dir = LOCAL_STORE_PATH
+
     metadata_file = dcd_mapping_dir / f"{scoreset_urn}_metadata.json"
     if not metadata_file.exists():
         url = f"{MAVEDB_BASE_URL}/api/v1/score-sets/{scoreset_urn}"
@@ -265,3 +269,27 @@ def get_scoreset_records(
                 raise ResourceAcquisitionError(msg) from e
 
     return _load_scoreset_records(scores_csv)
+
+
+def with_mavedb_score_set(fn: Callable) -> Callable:
+    @wraps(fn)
+    async def wrapper(*args, **kwargs) -> ScoresetMapping:  # noqa: ANN002
+        urn = args[0] if args else kwargs["urn"]
+        silent = kwargs.get("silent", False)
+
+        with tempfile.TemporaryDirectory(
+            prefix=f"{LOCAL_STORE_PATH.as_posix()}/"
+        ) as temp_dir:
+            # Set up metadata and scores for the current run. Now they will be accessible by these functions
+            # without the need to download the data again.
+            temp_dir_as_path = Path(temp_dir)
+            get_scoreset_metadata(urn, temp_dir_as_path)
+            get_scoreset_records(urn, silent, temp_dir_as_path)
+
+            # Pass the storage path of the temp directory to the wrapped function as a kwarg.
+            kwargs["store_path"] = temp_dir_as_path
+            v: ScoresetMapping = await fn(*args, **kwargs)
+
+            return v
+
+    return wrapper

src/dcd_mapping/schemas.py

@@ -21,9 +21,9 @@ class TargetSequenceType(str, Enum):
 class TargetType(str, Enum):
     """Define target gene types."""
 
-    PROTEIN_CODING = "Protein coding"
-    REGULATORY = "Regulatory"
-    OTHER_NC = "Other noncoding"
+    PROTEIN_CODING = "protein_coding"
+    REGULATORY = "regulatory"
+    OTHER_NC = "other_noncoding"
 
 
 class VrsVersion(str, Enum):

src/dcd_mapping/version.py

@@ -1,3 +1,3 @@
 """Provide dcd mapping version"""
 
-dcd_mapping_version = "2024.1.1"
+dcd_mapping_version = "2024.1.2"