Genomics researcher: nobody has 'solved' variant prioritization, but old-score-only views are outdated
anshulkundaje · x · 2026-09-18
- Anshul Kundaje argues variant prioritization is very hard: anyone claiming to have "solved" it is lying, but those insisting old scores and models are all we need are frozen in time.
- Aggregate scores (CADD, AVI, GPN-STAR) remain useful coupled with context-specific effects, but collapsing multi-dimensional scores into one universal number loses information.
- Non-coding variant effects are highly cell-context specific — a weakness of DNA LMs not conditioned on cell context; supervised sequence-to-omics models learn that context-specific regulatory logic.
- AI agents and model interpretation open the door to reasoning over multi-dimensional variant effect scores without lossy summarization.
More from Research
- Xiaomi MiMo's CodeSpecWeave trains coding agents on interdependent specs for long-horizon coding — XiaomiMiMo · 2026-11-02
- Mistral Large 4 writes a quantum chemistry SCF solver from scratch with one short prompt — MistralAI · 2026-11-02
- Hunting for the most concrete "curing cancer with AI" roadmap of 2025 — iskander · 2026-11-02
- New theory paper proves linear decay is the unique minimax-optimal stepsize schedule for SGD — konstmish · 2026-11-02
- Karpathy's nanochat: the best ChatGPT $100 can buy, near 60k GitHub stars — karpathy · 2026-11-02
- After reading 46 resources on agent evals, one author explains why her 95% score was misleading — blaizedsouza · 2026-11-02