AI and Automated Experiments Revolutionize Drug Design

Cambridge and Pfizer's 'Reactome' Platform Predicts Chemical Reactions, Speeding Up Pharmaceutical Development

Researchers from the University of Cambridge, in collaboration with Pfizer, have developed a platform that combines automated experiments with AI to predict how chemicals will react with one another, potentially accelerating the design process for new drugs.
The platform, called the 'reactome', uses a data-driven approach inspired by genomics and has been validated on a dataset of more than 39,000 pharmaceutically relevant reactions.
The 'reactome' approach identifies relevant correlations between reactants, reagents, and the performance of the reaction from the data, and also highlights gaps in the data itself.
In a related study, the team developed a machine learning model that enables chemists to introduce precise transformations to pre-specified regions of a molecule, facilitating faster drug design.
The researchers overcame the limitation of low data by pretraining the model on a large body of spectroscopic data, effectively teaching the model general chemistry, before fine-tuning it to predict intricate transformations.