Upload alphagenome_scorer.py with huggingface_hub

alphagenome_scorer.py

@@ -1,32 +1,28 @@
-"""GEMEO — AlphaGenome regulatory variant scorer (the non-coding piece).
+"""GEMEO — AlphaGenome regulatory/splice variant scorer (the non-coding piece).
 
 AlphaGenome (DeepMind, Nature 2026) is the regulatory-genomics SOTA — it scores
-how a variant affects gene expression, splicing, and chromatin, the NON-CODING
-effects that AlphaMissense (missense-only) and even Evo 2 capture less directly.
-This completes the ensemble: AlphaMissense (missense) + Evo 2 (coding/global) +
-AlphaGenome (regulatory).
+how a variant affects splicing, gene expression, and chromatin: the NON-CODING /
+splice effects that AlphaMissense (missense-only) and Evo 2 capture less directly.
+This completes the GEMEO genomic ensemble:
+    AlphaMissense (missense)  +  Evo 2 (coding/global)  +  AlphaGenome (splice/regulatory).
 
-Status (honest): the integration RUNS against the live API with a key — verified
-end-to-end (score_variant returns the real 19-modality effect outputs). What is
-NOT yet validated: a calibrated single-scalar pathogenicity. A naive max-|effect|
-aggregation across all modalities did NOT discriminate ClinVar splice variants
-(AUROC ~0.37 on n=28), so it is deliberately not used as a score. The correct path
-is AlphaGenome's recommended gene-centric variant scorers with per-modality
-aggregation (splicing scorer for splice variants, etc.) — that is the next step,
-and its AUROC will be committed to benchmark/results/ once it discriminates. Until
-then this module returns the raw effect summary, clearly labelled, not a
-pathogenicity probability — and never a fabricated number.
+Validated (real ClinVar splice variants, hg38, recommended SPLICE scorers,
+max|raw_score| aggregation): AUROC = 0.734 separating pathogenic vs benign splice
+variants (n=40); see benchmark/results/alphagenome_splice.json. This uses
+AlphaGenome's documented recommended variant scorers + tidy_scores aggregation —
+NOT a naive max over raw modality tensors (which did not discriminate).
 
-Requires an API key (free for research) in env var GEMEO_ALPHAGENOME_KEY, from
-https://deepmind.google.com/science/alphagenome . Without a key, returns None.
-
-    pip install alphagenome
-    export GEMEO_ALPHAGENOME_KEY=...
+Requires a free research API key in env var GEMEO_ALPHAGENOME_KEY
+(https://deepmind.google.com/science/alphagenome) and `pip install alphagenome`.
+Without a key, returns None — never a fabricated score.
 """
 from __future__ import annotations
 import os
 
 _KEY_ENV = "GEMEO_ALPHAGENOME_KEY"
+SPLICE_OUTPUTS = ("SPLICE_SITES", "SPLICE_SITE_USAGE", "SPLICE_JUNCTIONS")
+VALIDATION = {"task": "ClinVar splice (pathogenic vs benign)", "n": 40,
+              "aggregation": "max|raw_score| over recommended splice scorers", "auroc": 0.734}
 
 
 def available() -> bool:
@@ -39,34 +35,32 @@ def available() -> bool:
         return False
 
 
-def regulatory_effect(chrom: str, pos: int, ref: str, alt: str,
-                      window: int = 131072) -> dict | None:
-    """Score a variant's regulatory effect with AlphaGenome. Returns an effect
-    summary (max |delta| across expression/splice/chromatin tracks), or None if
-    the key/SDK is unavailable. No fabrication."""
+def splice_effect(chrom: str, pos: int, ref: str, alt: str,
+                  window: int = 2 ** 17) -> dict | None:
+    """AlphaGenome splice-disruption score for a variant (GRCh38 coords).
+    Returns max|raw_score| over the recommended splice scorers (gene-strand
+    matched), or None if the key/SDK is unavailable / no gene overlap.
+    Validated AUROC 0.734 on ClinVar splice variants."""
     key = os.environ.get(_KEY_ENV)
     if not key:
         return None
     try:
+        import numpy as np
         from alphagenome.data import genome
-        from alphagenome.models import dna_client
+        from alphagenome.models import dna_client, variant_scorers
         model = dna_client.create(key)
+        scorers = [variant_scorers.RECOMMENDED_VARIANT_SCORERS[k] for k in SPLICE_OUTPUTS]
         chrom = chrom if str(chrom).startswith("chr") else f"chr{chrom}"
         variant = genome.Variant(chromosome=chrom, position=int(pos),
                                   reference_bases=ref, alternate_bases=alt)
         interval = variant.reference_interval.resize(window)
-        scores = model.score_variant(interval=interval, variant=variant)
-        # aggregate to a single magnitude across modalities (recommended scorers)
-        mags = []
-        for s in scores:
-            try:
-                import numpy as np
-                mags.append(float(np.nanmax(np.abs(s.values))))
-            except Exception:
-                continue
-        return {"variant": f"{chrom}:{pos}{ref}>{alt}",
-                "regulatory_effect": max(mags) if mags else None,
-                "n_modalities": len(scores), "source": "alphagenome (live API)"}
+        out = model.score_variant(interval=interval, variant=variant, variant_scorers=scorers)
+        df = variant_scorers.tidy_scores(out)
+        if df is None or "raw_score" not in df:
+            return None
+        mag = float(np.nanmax(df["raw_score"].astype(float).abs().values))
+        return {"variant": f"{chrom}:{pos}{ref}>{alt}", "splice_disruption": round(mag, 4),
+                "source": "alphagenome (recommended splice scorers, live API)"}
     except Exception as e:  # pragma: no cover
         return {"error": str(e)[:120]}
 
@@ -74,7 +68,7 @@ def regulatory_effect(chrom: str, pos: int, ref: str, alt: str,
 if __name__ == "__main__":
     if not available():
         print("AlphaGenome scorer: NOT active — set GEMEO_ALPHAGENOME_KEY "
-              "(free, https://deepmind.google.com/science/alphagenome) and `pip install alphagenome`.")
+              "(free, https://deepmind.google.com/science/alphagenome) + `pip install alphagenome`.")
+        print(f"Validated when active: {VALIDATION}")
     else:
-        # FBN1 example (chr15, hg19) — only runs with a key
-        print(regulatory_effect("15", 48892414, "G", "A"))
+        print(splice_effect("7", 96184276, "C", "T"))  # example splice variant (hg38)