Learning to Make Generalizable and Diverse Predictions for Retrosynthesis
This work addresses the need for varied and robust reaction predictions in chemistry, though it is incremental as it builds on existing Transformer methods.
The authors tackled the problem of predicting diverse and generalizable chemical reactants for retrosynthesis by proposing a Transformer-based model with novel pre-training and a discrete latent variable, achieving improved performance and diversity on the USPTO-50k benchmark dataset.
We propose a new model for making generalizable and diverse retrosynthetic reaction predictions. Given a target compound, the task is to predict the likely chemical reactants to produce the target. This generative task can be framed as a sequence-to-sequence problem by using the SMILES representations of the molecules. Building on top of the popular Transformer architecture, we propose two novel pre-training methods that construct relevant auxiliary tasks (plausible reactions) for our problem. Furthermore, we incorporate a discrete latent variable model into the architecture to encourage the model to produce a diverse set of alternative predictions. On the 50k subset of reaction examples from the United States patent literature (USPTO-50k) benchmark dataset, our model greatly improves performance over the baseline, while also generating predictions that are more diverse.