Drug discovery chemists can cut wasted experiments by nearly a third

A new AI model can now predict which small changes to a drug molecule will have big effects on its potency. This means drug chemists can cut the number of early-stage experiments by nearly a third.

Drug discovery is a long, expensive process, often involving many trial-and-error experiments to find the right molecule. This new model helps chemists pinpoint the most important parts of a molecule to modify, avoiding many dead ends. It means they can find promising drug candidates faster and with less cost in the earliest stages.