Modern Hopfield Networks for Few- and Zero-Shot Reaction Template Prediction
This work addresses the challenge of finding synthesis routes for drug and material discovery, particularly for templates with few or zero training examples, representing an incremental improvement in template-based retrosynthesis methods.
The study tackled the problem of predicting reaction templates for single-step retrosynthesis in computer-assisted synthesis planning, using modern Hopfield networks to associate templates and molecules, resulting in improved predictive performance for top-k exact match accuracy for k≥5 on the USPTO-50k benchmark and faster inference speed.
Finding synthesis routes for molecules of interest is an essential step in the discovery of new drugs and materials. To find such routes, computer-assisted synthesis planning (CASP) methods are employed which rely on a model of chemical reactivity. In this study, we model single-step retrosynthesis in a template-based approach using modern Hopfield networks (MHNs). We adapt MHNs to associate different modalities, reaction templates and molecules, which allows the model to leverage structural information about reaction templates. This approach significantly improves the performance of template relevance prediction, especially for templates with few or zero training examples. With inference speed several times faster than that of baseline methods, we improve predictive performance for top-k exact match accuracy for $\mathrm{k}\geq5$ in the retrosynthesis benchmark USPTO-50k.