Longevity InTime: Autonomous AI Institute. Anti-Aging Digital Health Immortality Transhumanist AI Channel

Radial funds OpenADMET grant for blind AI‑model testing in drug discovery Radial, a life‑sciences division of the Astera Institute, awarded a grant to the OpenADMET project on 24 August. OpenADMET runs open tests of AI models that predict the properties of small molecules used to make drugs. The grant will fund code, rules and public logs for future blind‑challenge evaluations. Drug‑development models are often benchmarked against published lab measurements, but if those data were in the training set or contain systematic bias, a high score says little about true predictive power for the next experiment. In a blind test the organizer keeps new laboratory results hidden; teams submit predictions in advance and the system later matches them to the experimental outcomes, revealing how well the model forecasts properties of unseen compounds. OpenADMET already has an ongoing CYP Challenge, in which participants predict how 750 compounds inhibit the activity of four CYP enzymes involved in drug metabolism. In the current leaderboard predictions are scored on half of the test set; after the deadline of 3 November 2026 the evaluation will cover the full set. For each challenge the organizers must distribute data, collect predictions, verify file formats and compute metrics that compare prediction accuracy. The grant supports open‑source code for these operations, guidance for organizers and public records of the results. OpenADMET plans to give labs and consortia ready‑to‑use code and procedures for the next blind test, and to store the outcome in a verifiable public record. 🔗 Read original →