Mapping the Space of Chemical Reactions Using Attention-Based Neural Networks
This work provides a more efficient and accurate method for classifying organic reactions, which is beneficial for chemists and researchers navigating the complex chemical reaction space.
This paper addresses the tedious task of classifying organic reactions by demonstrating that transformer-based models can infer reaction classes from non-annotated, text-based representations of chemical reactions, achieving a classification accuracy of 98.2%. Additionally, the learned representations serve as reaction fingerprints that capture fine-grained differences between reaction classes more effectively than traditional methods.
Organic reactions are usually assigned to classes containing reactions with similar reagents and mechanisms. Reaction classes facilitate the communication of complex concepts and efficient navigation through chemical reaction space. However, the classification process is a tedious task. It requires the identification of the corresponding reaction class template via annotation of the number of molecules in the reactions, the reaction center, and the distinction between reactants and reagents. This work shows that transformer-based models can infer reaction classes from non-annotated, simple text-based representations of chemical reactions. Our best model reaches a classification accuracy of 98.2%. We also show that the learned representations can be used as reaction fingerprints that capture fine-grained differences between reaction classes better than traditional reaction fingerprints. The insights into chemical reaction space enabled by our learned fingerprints are illustrated by an interactive reaction atlas providing visual clustering and similarity searching.