Inductive has launched Beacon-2, a system that predicts the efficacious human dose of a small molecule directly from its chemical structure, the company said. Dose determines whether a compound can become a drug, and reflects both a molecule's potency and what the body does to it. A drug that works at a lower dose also carries less risk of off-target toxicity, the company said.
Beacon-2 predicts absorption, distribution, metabolism, excretion and toxicity, estimates potency, and combines those outputs through mechanistic models of pharmacokinetics. Inductive said the ADMET models underneath the system have won three consecutive OpenADMET blind challenges, beating more than 750 competitors from AI and pharmaceutical companies.
A technical post published today covers how Beacon-2 works and its projections across 20 real-world drug programs and 325 publicly available compounds from the ExpansionRx OpenADMET competition. To test how the system affects agentic design workflows, Inductive gave its medicinal chemistry agent Indy a recently disclosed SARS-CoV-2 compound to optimize. Across five autonomous design cycles, Indy improved the predicted human dose 17-fold, which the company said can be the difference between a drug program being killed and one reaching clinical trials.
Josh Haimson, co-founder and chief executive of Inductive, said chemists have always had to balance a dozen separate properties against each other when deciding whether a compound is worth making, and that dose is what those properties add up to. He said computing dose from structure changes which compounds a team decides to make and helps teams identify the highest quality compounds faster. Beacon-2 is available to partners through Compass, where the company said it is already running on live drug discovery programs.
