MIT researchers have developed a new generative AI approach called FlowER (Flow matching for Electron Redistribution) to predict chemical reactions with improved accuracy and reliability. This system incorporates fundamental physical principles, such as the laws of conservation of mass, to ensure that predictions are physically realistic.
FlowER explicitly tracks all electrons in a reaction, preventing spurious additions or deletions. The system uses a matrix to represent electrons in a reaction, with nonzero values for bonds or lone electron pairs and zeros for a lack thereof. This representation helps conserve both atoms and electrons simultaneously.
The FlowER model has shown promising results, matching or outperforming existing approaches in finding standard mechanistic pathways. Additionally, it can generalize to previously unseen reaction types, making it a valuable tool for predicting reactions in various fields, including medicinal chemistry, materials discovery, combustion, atmospheric chemistry, and electrochemical systems.
The model's developers plan to expand its understanding of metals and catalytic cycles. The FlowER model is available on GitHub, allowing researchers to access and build upon the work. This innovative approach has the potential to accelerate the discovery of new chemical reactions and improve our understanding of complex chemical processes.
