AI model predicts and explains cell-type-specific TSS usage driven by MEF2 motif
anshulkundaje · x · 2026-11-07
The thread's capstone: the model predicts — and explains — cell-type-specific transcription start site (TSS) usage. The attached figure shows one enhancer where the cardiac/muscle-specific MEF2 TF motif drives usage of a downstream TSS, while a different TSS is active in ileum.
- Observed vs predicted PRO-cap coverage closely match in both cell types
- DeepLIFT attribution clearly recovers Inr, MEF2, AP-1, ETS, ZSCAN motifs
- Demonstrates mechanistic interpretability, not just prediction
More from Research
- METR data implies AI task time horizons now double every ~2 months — scaling01 · 2026-11-07
- Open-source EuroSafeGuard outdetects 16 guard models on jailbreaks, halving PAIR attack success — ZhijingJin · 2026-11-07
- How EuroSafeGuard works: activation probes triage traffic before heavy classifiers — ZhijingJin · 2026-11-07
- EuroSafeGuard: first fully open Constitutional Classifiers guardrail beats 16 open guards — ZhijingJin · 2026-11-07
- OpenVax revamps personalized cancer vaccine toolkit; OpenVax2 adds long-read support — iskander · 2026-11-07
- Google's Muir linear attention fits 10M tokens in a 25MB state, beats GDN-2 — CategoryNormal149 · 2026-11-07