AI Drug Discovery Reaches Clinical Proof Point
Isomorphic Labs, the drug discovery spinout from Google DeepMind, announced this week that its first independently designed drug candidate — ISL-001, an AI-designed kinase inhibitor targeting ANCA-associated vasculitis — met its primary endpoint in a Phase II clinical trial, showing statistically significant reduction in disease activity versus placebo.
What the Result Means
Phase II trials test efficacy and dose-ranging in a relatively small patient population. ISL-001's success does not guarantee Phase III success (most drugs fail in Phase III), but it represents the first published Phase II win for a drug discovered entirely through AI-guided structure-based design without the traditional medicinal chemistry iteration cycle.
The Technology Behind It
ISL-001 was designed using AlphaFold-derived protein structure predictions to identify a binding site on the target kinase, followed by generative molecular design to create candidate compounds with high predicted affinity and selectivity. The entire hit-to-lead optimization process took 6 months, versus a typical 3-5 years for conventional approaches.
Partnership Implications
Isomorphic Labs has partnership agreements with Eli Lilly and Novartis that include milestone payments triggered by clinical trial success. This Phase II result is expected to trigger a $45 million milestone payment from Eli Lilly and advances ISL-001 toward an anticipated Phase III start in Q1 2027.
Industry Context
Multiple AI drug discovery companies including Recursion Pharmaceuticals, Exscientia (acquired by Recursion), and Insilico Medicine have molecules in clinical trials. The field is approaching its first Phase III results from fully AI-designed drug candidates.