claim
active
claim:evee-outperforms-existing-methods-for-variant-pathogenicity-predictionEvee outperforms existing methods for variant pathogenicity prediction
Interpretive claim supported by the high AUROC findings.
Source paper
extracted_from(2026) · Pearce, Michael · Dooms, Thomas · Yamamoto, Ryo · Meehl, Joshua +18
Neighborhood — ranked by edge-count
Papers (1)
paper
Findings (2)
finding
- EVEE achieves state-of-the-art performance on variant pathogenicity classification, outperforming existing methods.
- EVEE demonstrates strong generalization to indels without explicit training, indicating learned mechanistic principles.
Communities (3)
community
- Spans attention head decomposition, benchmark awareness, and genomic pathogenicity prediction via neural models.
- Using genomic foundation model internals to generate disruption profiles that explain variant effects mechanistically, achieving 0.997 AUROC on ClinVar pathogenicity prediction.
- EVEE tool predicts and mechanistically explains effects of 4.2M human genome variants using model internals.
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- The method that predicts and explains variant pathogenicity using Evo 2, producing disruption profiles.
- Core interpretability claim distinguishing EVEE from black-box prediction tools; applies interpretability for science.
- Core task of predicting whether genetic variants cause disease, central to clinical genomics.
- Scale claim, demonstrating whole-genome applicability.
- The property of a genetic variant being disease-causing or benign; the main prediction target.
- Claim by Comolatti & Hoel (2022) endorsed by this survey.
- Surprising finding that the two evaluation methods diverge in their relationship with persistence