An AI that speaks 'cell' just found a drug pair that fights prostate cancer together

Drug combinations are cancer's puzzle: thousands of possible pairings, far too many to test one by one in dishes. A UCLA-led team built an AI platform that reads how drugs act on cells at the level of their inner signals — 'mechanism-informed' language models — and used it, together with realistic 3D tumour screening, to nominate pairings worth a real test. Its pick: berberine, a plant-derived compound, combined with enzalutamide, an established prostate cancer drug. In prostate cancer models the pair worked in synergy — better together than either alone. The study, in Advanced Science, is a proof of method as much as of medicine: a way to search the vast space of combinations with a map instead of luck. 'With appropriate scientific design and validation, they can support drug discovery and generate hypotheses that can be tested experimentally,' the authors write. Human trials remain ahead — but the shortlist just got much smarter.

SDG 3 · Good health and well-being — Mechanism: tech innovation