pheval-exomiser 0.4.5__tar.gz → 0.4.8__tar.gz

{pheval_exomiser-0.4.5 → pheval_exomiser-0.4.8}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pheval_exomiser
-Version: 0.4.5
+Version: 0.4.8
 Summary: 
 Author: Yasemin Bridges
 Author-email: y.bridges@qmul.ac.uk
@@ -17,7 +17,7 @@ Requires-Dist: google (>=3.0.0,<4.0.0)
 Requires-Dist: numpy (<2)
 Requires-Dist: oaklib (>=0.5.12,<0.6.0)
 Requires-Dist: phenopackets (>=2.0.2,<3.0.0)
-Requires-Dist: pheval (>=0.6.7,<0.7.0)
+Requires-Dist: pheval (>=0.7.0,<0.8.0)
 Requires-Dist: pyaml (>=21.10.1,<22.0.0)
 Requires-Dist: pydantic (>=2.7.1,<3.0.0)
 Description-Content-Type: text/markdown

{pheval_exomiser-0.4.5 → pheval_exomiser-0.4.8}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "pheval_exomiser"
-version = "0.4.5"
+version = "0.4.8"
 description = ""
 authors = ["Yasemin Bridges <y.bridges@qmul.ac.uk>",
     "Julius Jacobsen <j.jacobsen@qmul.ac.uk>",
@@ -19,7 +19,7 @@ oaklib = "^0.5.12"
 docker = "^6.0.1"
 pydantic = "^2.7.1"
 numpy = "<2"
-pheval = "^0.6.7"
+pheval = "^0.7.0"
 
 [tool.poetry.dev-dependencies]
 pytest = "^7.1.2"

{pheval_exomiser-0.4.5 → pheval_exomiser-0.4.8}/src/pheval_exomiser/post_process/post_process_results_format.py

@@ -13,6 +13,19 @@ from pheval.post_processing.post_processing import (
 )
 from pheval.utils.file_utils import files_with_suffix
 
+EXOMISER_LT_15 = {"combinedScore", "priorityScore", "variantScore", "pValue"}
+EXOMISER_GTE_15 = {"geneCombinedScore", "geneVariantScore", "pValue"}
+
+ALL_SCORE_NAMES = sorted(EXOMISER_LT_15 | EXOMISER_GTE_15)
+
+
+def _allowed_score_names(exomiser_version: str) -> set[str]:
+    return (
+        EXOMISER_GTE_15
+        if version.parse(exomiser_version) >= version.parse("15.0.0")
+        else EXOMISER_LT_15
+    )
+
 
 class ModeOfInheritance(Enum):
     AUTOSOMAL_DOMINANT = 1
@@ -27,6 +40,28 @@ class ModeOfInheritance(Enum):
     MT = 3
 
 
+def check_score_name(score_name: str, version: str):
+    """
+    Validates the provided score name for compatibility with the specified Exomiser version.
+
+    Args:
+        score_name: str
+            The name of the score to validate.
+        version: str
+            The Exomiser version for which the validation is performed.
+    Raises:
+        click.BadParameter: Raised if the provided score name is not valid,
+        including the list of allowed score names in the error message.
+    """
+    allowed_score_names = _allowed_score_names(version)
+    if score_name not in allowed_score_names:
+        raise click.BadParameter(
+            f"'{score_name}' is not valid for Exomiser {version}. "
+            f"Allowed values: {', '.join(sorted(allowed_score_names))}.",
+            param_hint="--score-name",
+        )
+
+
 def trim_exomiser_result_filename(exomiser_result_path: Path) -> Path:
     """Trim suffix appended to Exomiser JSON result path."""
     return Path(str(exomiser_result_path.name).replace("-exomiser", ""))
@@ -46,14 +81,21 @@ def extract_gene_results_from_json(
 
 
 def extract_gene_results_from_parquet(
-    exomiser_parquet_result: pl.DataFrame, score_name: str
+    exomiser_parquet_result: pl.DataFrame, score_name: str, variant_analysis: bool
 ) -> pl.DataFrame:
+    if variant_analysis:
+        exomiser_parquet_result = exomiser_parquet_result.filter(
+            pl.col("isContributingVariant") == True  # noqa
+        )
     return exomiser_parquet_result.select(
         [
             pl.col("geneSymbol").alias("gene_symbol"),
             pl.col("ensemblGeneId").alias("gene_identifier"),
             pl.col(score_name).fill_null(0).round(4).alias("score"),
         ]
+    ).unique(
+        subset=["gene_symbol", "gene_identifier"],
+        keep="first",
     )
 
 
@@ -241,7 +283,7 @@ def create_standardised_results(
             exomiser_result = pl.read_json(exomiser_result_path, infer_schema_length=None)
         if gene_analysis:
             gene_results = (
-                extract_gene_results_from_parquet(exomiser_result, score_name)
+                extract_gene_results_from_parquet(exomiser_result, score_name, variant_analysis)
                 if use_parquet
                 else extract_gene_results_from_json(exomiser_result, score_name)
             )
@@ -308,11 +350,9 @@ def create_standardised_results(
 @click.option(
     "--score-name",
     "-s",
+    type=click.Choice(ALL_SCORE_NAMES),
+    help="Score column to extract (valid values depend on Exomiser version).",
     required=True,
-    help="Score name to extract from results.",
-    type=click.Choice(["combinedScore", "priorityScore", "variantScore", "pValue"]),
-    default="combinedScore",
-    show_default=True,
 )
 @click.option(
     "--sort-order",
@@ -376,6 +416,7 @@ def post_process_exomiser_results(
         if disease_analysis
         else None
     )
+    check_score_name(score_name, version)
     create_standardised_results(
         result_dir=results_dir,
         output_dir=output_dir,