MARS: A Motif-based Autoregressive Model for Retrosynthesis Prediction
Abstract: Retrosynthesis is a critical task in drug discovery, aimed at finding a viable pathway for synthesizing a given target molecule. Many existing approaches frame this task as a graph-generating problem. Specifically, these methods firstly identify the reaction center, and break a targeted molecule accordingly to generate synthons. Reactants are generated by either adding atoms sequentially to synthon graphs or directly adding proper leaving groups. However, both of these strategies have limitations. Adding atoms results in a long prediction sequence which increases the complexity of generation, while adding leaving groups can only consider those in the training set which results in poor generalization. In this paper, we propose a novel end-to-end graph generation model for retrosynthesis prediction, which sequentially identifies the reaction center, generates the synthons, and adds motifs to the synthons to generate reactants. Since chemically meaningful motifs are bigger than atoms and smaller than leaving groups, our method enjoys lower prediction complexity than adding atoms and better generalization than adding leaving groups. We evaluate our proposed model on a benchmark dataset and show that it significantly outperforms previous state-of-the-art models. Furthermore, we conduct an ablation study to investigate the contribution of each component of our proposed model to the overall performance on the benchmark dataset. Our results demonstrate the effectiveness of our model in predicting rethresynthesis pathways and suggest its potential as a valuable tool in drug discovery.
Submission Length: Long submission (more than 12 pages of main content)
Assigned Action Editor: ~Lei_Li11
Submission Number: 1340
Loading