Kemal Yaylali
e76ae847a1
A review of the ranking's arithmetic found four things wrong, all of which made the score look better informed than it was. Measurements below are from this repo, not estimates. **Components now abstain instead of inventing a number.** A run without a VEP cache returns no allele frequencies, and rarity_score(None) read that as "absent from gnomAD, therefore maximally rare" and awarded every variant a free 0.25. jobs.has_frequencies / has_effect_scores record what the run actually produced, absent components are dropped from the weighted mean, and the remaining weights are renormalised so the score keeps its meaning. The UI shows "not looked up" rather than a bar, and the funnel stops calling a step "rare" when nothing was filtered. **Allele frequency is no longer a model feature.** It dominated: the same missense variant scored 0.887 at AF 0 and 0.0003 at AF 0.01. That double- counted, because the ranking already scores frequency explicitly, putting ~45% of every rank on one measurement; and it was circular, because ACMG assigns ClinVar's benign labels using frequency (BA1/BS1). Retraining without it moves missense AUROC from 0.872 to 0.500 — exactly random. The old figure was allele frequency, not variant-effect knowledge. The model therefore abstains unless CADD or AlphaMissense is present, since otherwise it only restates the consequence class. **Phenotype matching is weighted by information content** and HPO annotations are propagated up the ontology. Counting terms alike let "global developmental delay" (IC 0.93) count as much as "dilated left subclavian artery" (IC 7.88). **A real bug in the propagation, found by checking it.** The ancestor walk read a pre-order DFS backwards, which on a DAG lets a term resolve before one of its parents and inherit that parent alone instead of its lineage. It dropped 399 terms out of the phenotype branch, Camptodactyly and Chiari malformation among them. Now a true post-order, tested against a reference transitive closure. The ontology arithmetic moved to rarelens_ml.hpo so it is covered by tests, and rarelens_ml.benchmark measures the whole thing: across 10,178 published cases the causal gene ranks first 45.9-81.0% of the time against 5,269 genes, versus 0.02% for chance. docs/data.md reports that with its contamination (HPO's annotations come from these same case reports), and includes the measurement showing information-content weighting earns its place while propagation does not - kept anyway, for a reason the docs argue rather than assume.
106 lines
3.8 KiB
Python
106 lines
3.8 KiB
Python
import uuid
|
|
|
|
import numpy as np
|
|
import pandas as pd
|
|
import pytest
|
|
from factories import seed_case
|
|
from httpx import AsyncClient
|
|
from sqlalchemy import select
|
|
|
|
from app.db import SessionLocal
|
|
from app.models import JobStatus, Prediction, Variant
|
|
from app.services import scoring
|
|
|
|
|
|
def variant(**kw: object) -> Variant:
|
|
fields: dict = {"chrom": "22", "pos": 1, "ref": "A", "alt": "G", "annotations": {}}
|
|
fields.update(kw)
|
|
return Variant(**fields)
|
|
|
|
|
|
def test_raw_frame_sends_the_model_contract_columns() -> None:
|
|
frame = scoring.raw_frame([
|
|
variant(impact="HIGH", consequence="stop_gained", gnomad_af=None,
|
|
annotations={"CADD_PHRED": "35", "am_pathogenicity": "0.98"}),
|
|
variant(impact="LOW", consequence="synonymous_variant", gnomad_af=0.2, annotations={}),
|
|
])
|
|
assert list(frame.columns) == scoring.RAW_COLUMNS
|
|
assert frame["impact"].tolist() == ["HIGH", "LOW"]
|
|
assert frame["cadd_phred"].iloc[0] == "35"
|
|
assert pd.isna(frame["cadd_phred"].iloc[1])
|
|
# Frequency is scored by the ranking, auditably; sending it here too counted it twice.
|
|
assert "gnomad_af" not in frame.columns
|
|
|
|
|
|
class FakeModel:
|
|
def __init__(self, score: float) -> None:
|
|
self.score = score
|
|
|
|
def predict(self, frame: pd.DataFrame) -> np.ndarray:
|
|
assert list(frame.columns) == scoring.RAW_COLUMNS
|
|
return np.full(len(frame), self.score)
|
|
|
|
|
|
async def make_case(status: JobStatus, n_variants: int) -> tuple[uuid.UUID, uuid.UUID]:
|
|
return await seed_case(
|
|
status=status,
|
|
variants=[{"pos": i + 1, "annotations": {}} for i in range(n_variants)],
|
|
)
|
|
|
|
|
|
async def predictions(job_id: uuid.UUID) -> list[Prediction]:
|
|
async with SessionLocal() as s:
|
|
rows = await s.scalars(select(Prediction).join(Variant).where(Variant.job_id == job_id))
|
|
return list(rows)
|
|
|
|
|
|
@pytest.mark.usefixtures("db")
|
|
async def test_scoring_twice_updates_instead_of_failing(
|
|
client: AsyncClient, monkeypatch: pytest.MonkeyPatch
|
|
) -> None:
|
|
case_id, job_id = await make_case(JobStatus.succeeded, n_variants=3)
|
|
|
|
monkeypatch.setattr(scoring, "load_model", lambda: (FakeModel(0.9), "7"))
|
|
r = await client.post(f"/api/cases/{case_id}/score")
|
|
assert r.status_code == 200, r.text
|
|
assert r.json() == {"case_id": str(case_id), "scored": 3, "model_version": "7"}
|
|
|
|
monkeypatch.setattr(scoring, "load_model", lambda: (FakeModel(0.2), "8"))
|
|
r = await client.post(f"/api/cases/{case_id}/score")
|
|
assert r.status_code == 200, r.text
|
|
|
|
preds = await predictions(job_id)
|
|
assert len(preds) == 3
|
|
assert {(p.score, p.model_version) for p in preds} == {(0.2, "8")}
|
|
|
|
|
|
@pytest.mark.usefixtures("db")
|
|
async def test_scoring_an_unknown_case_is_404(client: AsyncClient) -> None:
|
|
r = await client.post(f"/api/cases/{uuid.uuid4()}/score")
|
|
assert r.status_code == 404
|
|
|
|
|
|
@pytest.mark.usefixtures("db")
|
|
async def test_scoring_before_the_annotation_finishes_is_409(client: AsyncClient) -> None:
|
|
case_id, _ = await make_case(JobStatus.running, n_variants=1)
|
|
r = await client.post(f"/api/cases/{case_id}/score")
|
|
assert r.status_code == 409
|
|
|
|
|
|
@pytest.mark.usefixtures("db")
|
|
async def test_an_unreachable_model_registry_is_explained_not_a_500(
|
|
client: AsyncClient, monkeypatch: pytest.MonkeyPatch
|
|
) -> None:
|
|
case_id, _ = await make_case(JobStatus.succeeded, n_variants=1)
|
|
|
|
def unreachable() -> tuple:
|
|
raise ConnectionError("connection refused to http://localhost:5000")
|
|
|
|
monkeypatch.setattr(scoring, "load_model", unreachable)
|
|
r = await client.post(f"/api/cases/{case_id}/score")
|
|
assert r.status_code == 503
|
|
detail = r.json()["detail"]
|
|
# Names what is missing; the exception itself belongs in the server log, not the UI.
|
|
assert "rarelens-pathogenicity@production" in detail
|
|
assert "Max retries" not in detail and "Traceback" not in detail
|