Modelling Chemical Reasoning to Predict Reactions

TL;DR

This paper introduces a data-driven model that mimics chemical reasoning by predicting reactions through a large knowledge graph, outperforming rule-based systems and enabling rapid, high-throughput reaction hypothesis generation.

Contribution

The authors constructed a comprehensive chemical knowledge graph and developed a model that generalizes beyond known reactions to predict and discover novel chemical transformations.

Findings

Outperforms rule-based expert systems in reaction prediction

Generalizes beyond known reaction types, including transition-metal catalysis

Predicts reactions in sub-second time, enabling high-throughput hypothesis generation

Abstract

The ability to reason beyond established knowledge allows Organic Chemists to solve synthetic problems and to invent novel transformations. Here, we propose a model which mimics chemical reasoning and formalises reaction prediction as finding missing links in a knowledge graph. We have constructed a knowledge graph containing 14.4 million molecules and 8.2 million binary reactions, which represents the bulk of all chemical reactions ever published in the scientific literature. Our model outperforms a rule-based expert system in the reaction prediction task for 180,000 randomly selected binary reactions. We show that our data-driven model generalises even beyond known reaction types, and is thus capable of effectively (re-) discovering novel transformations (even including transition-metal catalysed reactions). Our model enables computers to infer hypotheses about reactivity and reactions by only considering the intrinsic local structure of the graph, and because each single reaction prediction is typically achieved in a sub-second time frame, our model can be used as a high-throughput generator of reaction hypotheses for reaction discovery.