Cell Type launches CT-Perturb, a multimodal language model that predicts how cells respond to drugs and gene knockouts
david_van_dijk · x · 2026-10-06
Cell Type introduced CT-Perturb, a multimodal language model built over tokenized biology that predicts how cells respond to perturbations—without the lab. It handles compounds, gene knockouts, cytokines, and their combinations, aiming to pre-screen experiments computationally before wet-lab work. The team shared the model, a technical explainer thread, and the results behind the "screening beyond the lab" claim—a notable AI-for-Science release.
More from Research
- 'Chicken' can make base models reason: right first tokens match RL training — phillip_isola · 2026-10-06
- 3SUM and APSP hypotheses refuted as subquadratic breakthrough lands — aran_nayebi · 2026-10-06
- 18,000+ runs: new study shows how agent configuration shapes performance on 4 scientific tasks — zeynepakata · 2026-10-06
- Hamel Husain: similarity metrics like ROUGE don't work for LLM output evals — HamelHusain · 2026-10-06
- Nearly half of DOL-recognized jobs have zero agentic AI tool coverage, Cohere finds — Cohere_Labs · 2026-10-06
- Princeton-affiliated paper: stochastic contexts make adversarial online learning efficiently computable — HazanPrinceton · 2026-10-06