Leveraging Reaction-aware Substructures for Retrosynthesis Analysis

TL;DR

This paper introduces a reaction-aware substructure decoding model for retrosynthesis analysis, improving prediction accuracy and providing better insights by focusing on stable substructures during chemical reactions.

Contribution

The paper presents a novel substructure-level decoding approach that automatically extracts reaction-aware substructures, enhancing retrosynthesis prediction performance.

Findings

Improved accuracy over previous models.

Substructure extraction accuracy correlates with performance.

Provides better insights for decision-making.

Abstract

Retrosynthesis analysis is a critical task in organic chemistry central to many important industries. Previously, various machine learning approaches have achieved promising results on this task by representing output molecules as strings and autoregressively decoded token-by-token with generative models. Text generation or machine translation models in natural language processing were frequently utilized approaches. The token-by-token decoding approach is not intuitive from a chemistry perspective because some substructures are relatively stable and remain unchanged during reactions. In this paper, we propose a substructure-level decoding model, where the substructures are reaction-aware and can be automatically extracted with a fully data-driven approach. Our approach achieved improvement over previously reported models, and we find that the performance can be further boosted if the accuracy of substructure extraction is improved. The substructures extracted by our approach can provide users with better insights for decision-making compared to existing methods. We hope this work will generate interest in this fast growing and highly interdisciplinary area on retrosynthesis prediction and other related topics.