AI drug discovery
Generative Design

At Exscientia evolutionary algorithms are applied routinely to invent new chemical matter. These systems are an embodiment of Generative Design and can be applied in both 2D and 3D space, depending on both the task at hand and information available, such as whether protein structure information is known.

Starting with seed data specific to a project, novel project-relevant compounds (sometimes many millions) are evolved autonomously and selection pressure is applied to ensure that only those predicted to be performant (an appropriate balance of potency, selectivity and DMPK properties) are made available to the Active Learning systems for selection and submission for synthesis and testing.

Through progressive rounds of Automated Invention that are tightly aligned with experimental synthesis and testing of prioritised molecules, a rapid learning towards key projects goals is enabled enabling sector beating performance in progressing from Hits through Leads to Candidate.