Reaction Graph: Towards Reaction-Level Modeling for Chemical Reactions with 3D Structures

Yingzhao Jian, Yue Zhang, Ying Wei, Hehe Fan, Yi Yang

TL;DR: In this paper, we introduce Reaction Graph (RG), a unified graph representation that encapsulates the 3D structures within chemical reactions.

Abstract: Accurately modeling chemical reactions using Artificial Intelligence (AI) can accelerate discovery and development, especially in fields like drug design and material science. Although AI has made remarkable advancements in single molecule recognition, such as predicting molecular properties, the study of interactions between molecules, particularly chemical reactions, has been relatively overlooked. In this paper, we introduce Reaction Graph (RG), a unified graph representation that encapsulates the 3D molecular structures within chemical reactions. RG integrates the molecular graphs of reactants and products into a cohesive framework, effectively capturing the interatomic relationships pertinent to the reaction process. Additionally, it incorporates the 3D structure information of molecules in a simple yet effective manner. We conduct experiments on a range of tasks, including chemical reaction classification, condition prediction, and yield prediction. RG achieves the highest accuracy across six datasets, demonstrating its effectiveness. The code is available at https://github.com/Shadow-Dream/Reaction-Graph.

Lay Summary:

Chemical reactions happen when atoms rearrange themselves, turning molecules into something new, like making medicines or new materials. However, current artificial intelligence (AI) methods still are not great at guessing how reactions will turn out, including conditions, yields, or reaction types.
We think this is because these AI methods can't easily see how atoms move around or clearly understand molecules' 3D shapes.

To fix that, we made something called the Reaction Graph (RG). RG connects each atom before and after a reaction, clearly showing how they rearrange. It also highlights simple triangles to show the molecules' 3D shapes, which surprisingly helps a lot.

When we tested RG, it consistently performed better than other methods at predicting reaction conditions, yields, and types. RG can potentially help chemists do better research and make better products.

Link To Code: https://github.com/Shadow-Dream/Reaction-Graph

Primary Area: Deep Learning->Other Representation Learning

Keywords: Chemical Reaction Modeling, Reaction Graph, 3D Molecular Structure, AI for Chemistry

Submission Number: 6107