Chemical engineers at Columbia have trained a neural network model for retrosynthesis in organic chemistry.
Retrosynthesis is used by organic chemists to work backwards from the target molecule of a chemical reaction, to the starting materials. Columbia University School of Engineering and Applied Science has developed a deep learning technique for retrosynthetic analysis.
According to Columbia University School of Engineering and Applied Science, the chemist faces many complex, interrelated questions, such as: “Of the tens of thousands of different chemical reactions, which one should you choose to create the target molecule? Once that decision is made, you may find yourself with multiple reactant molecules needed for the reaction. If these molecules are not available to purchase, then how do you select the appropriate reactions to produce them?”
For this reason it is important for chemists to be able to make intelligent decisions to navigate the possible paths in retrosynthetic analysis.
The AI neural network model gives researchers a framework for designing chemical syntheses in line with specified objectives such as synthesis cost, safety, and sustainability.
Will the neural network model outperform humans?
Alán Aspuru-Guzik, professor of chemistry and computer science at the University of Toronto, who was not involved with the study. Aspuru-Guzik said: “Reinforcement learning has created computer players that are much better than humans at playing complex video games. Perhaps retrosynthesis is no different! This study gives us hope that reinforcement-learning algorithms will be perhaps one day better than human players at the ‘game’ of retrosynthesis.”
Bishop adds: “We expect that our retrosynthesis game will soon follow the way of chess and Go, in which self-taught algorithms consistently outperform human experts. And we welcome competition. As with chess-playing computer programs, competition is the engine for improvements in the state-of-the-art, and we hope that others can build on our work to demonstrate even better performance.”