NetGoodIndexSubmit a correction

Benefit Ledger · verified · Biology

AlphaGenome predicts multimodal regulatory effects of DNA variants

DeepMind’s AlphaGenome takes up to 1 Mb of DNA and predicts gene expression, splicing, chromatin, and contact tracks, matching or beating external models on 25 of 26 variant-effect tests in a 28 January 2026 Nature paper.

28 Jan 2026Tier 3 MajorMethodology 0.1

Current score

+1.98

10 base · Major (tier 3 of 5, 10 pts)
× 0.9000 attribution · Primary causal contribution
× 0.7000 evidence · Peer review or independent validation
× 0.4500 realization · Experimentally validated
× 0.7000 durability
Event-level product before credit split: 1.98

Unified long-context regulatory model is a major methods advance (tier 3). High attribution. Nature peer review. Realization is API/research, not clinical care. Durable as a published method.

What happened

Announced in 2025 and peer-reviewed in 2026, AlphaGenome unifies long-context sequence input with base-resolution multimodal outputs. The authors showed it can recover mechanisms of known TAL1 oncogene variants and released APIs for research. It is not a clinical genome interpreter. Realization is a research tool plus benchmarks, not diagnostic use.

Model attribution

+1.98

AlphaGenome

Sequence model for genome tracks and variant-effect scores.

The published model is the result.

Attribution 0.9000 · Credit share 100% · Google DeepMind

Claims

  • AlphaGenome matched or exceeded the strongest external models on 25 of 26 variant-effect evaluations in the Nature paper.

    outcome · supported

Sources

primary sources

Secondary domains: Medicine & Health

Revision history

  • 13 Sep 2026 · 0.00 → 1.98

    Initial adjudicated seed score under methodology 0.1.