AI-designed drug candidate reverses biological-age markers in a phase 2a trial
Insilico's rentosertib — its target and molecule designed with AI — showed shifts toward younger predicted biological ages across six aging clocks in treated patients.
A study published 7 September in Nature Biotechnology revisited data from a phase 2a trial of rentosertib, an experimental drug from Insilico Medicine developed with AI to treat idiopathic pulmonary fibrosis. AI identified the target (a protein called TNIK) and designed the molecule that inhibits it.
Researchers applied six “aging clocks” — computational models that estimate biological age from blood proteins — to 42 participants over 12 weeks. All six detected shifts toward younger predicted ages among treated patients, while placebo recipients showed little change or slight increases. At week four, patients on 60mg daily showed reductions of roughly 2.7–3.5 years on four chronological-age clocks.
Why this scores low (good news)
The researchers are explicit about the limits: these are model predictions, and they cannot yet fully separate effects on aging from improvements in lung disease. But it is a concrete, peer-reviewed signal that AI-designed molecules can move real clinical endpoints — the kind of event that pulls the ARQ down.
Risk 0 · Mitigation 4.