Mosaic model engineers cell states to expose hidden disease variants, opening chromatin at 28% of AMD loci vs 5-7% at rest
anshulkundaje · x · 2026-10-14
A team led by @thenormanlab uses TF perturbation and a computational model called Mosaic to engineer cell states that reveal phenotypes normally silent in standard cell models. Key findings: TF activation opens chromatin at 28% of fine-mapped AMD variants vs 5-7% at rest; omnigenic convergence onto Mendelian retinal disease genes is state-dependent; and chromatin priming—where one TF opens binding sites for another—predicts which TF pairs create new states, letting Mosaic predict novel disease gene regulators.
More from Research
- Google researcher: general capabilities plus tool use are displacing domain-specific medical AI — davidstutz92 · 2026-10-14
- Google's AMIE matched doctor diagnoses in 90% of 100 real-clinic patient interactions with zero safety stops — davidstutz92 · 2026-10-14
- SREGym team presents at Frontier Data Summit and teases SREGym 2.0 — tianyin_xu · 2026-10-14
- New paper claims OpenAI's Navier-Stokes proof and its Lean code don't match — peclo · 2026-10-14
- Submarine cyber ops study: operator cognitive capacity swings with the tactical environment — J_MaestreVidal · 2026-10-14
- Recova Trains Robots to Self-Recover in Simulation, Cutting Human Intervention From 87.5% to 0% — Lianhuiq · 2026-10-14