Molecule Edit Graph Attention Network: Modeling Chemical Reactions as Sequences of Graph Edits

TL;DR

MEGAN is a novel neural network model that predicts chemical reactions as sequences of graph edits, enabling efficient exploration of reaction space and achieving state-of-the-art accuracy in retrosynthesis tasks.

Contribution

The paper introduces MEGAN, a new end-to-end model that represents chemical reactions as sequences of graph edits, improving retrosynthesis prediction accuracy and scalability.

Findings

Achieves state-of-the-art accuracy on standard benchmarks.

Effectively models chemical reactions as sequences of graph edits.

Scales to large datasets for retrosynthesis prediction.

Abstract

The central challenge in automated synthesis planning is to be able to generate and predict outcomes of a diverse set of chemical reactions. In particular, in many cases, the most likely synthesis pathway cannot be applied due to additional constraints, which requires proposing alternative chemical reactions. With this in mind, we present Molecule Edit Graph Attention Network (MEGAN), an end-to-end encoder-decoder neural model. MEGAN is inspired by models that express a chemical reaction as a sequence of graph edits, akin to the arrow pushing formalism. We extend this model to retrosynthesis prediction (predicting substrates given the product of a chemical reaction) and scale it up to large datasets. We argue that representing the reaction as a sequence of edits enables MEGAN to efficiently explore the space of plausible chemical reactions, maintaining the flexibility of modeling the reaction in an end-to-end fashion, and achieving state-of-the-art accuracy in standard benchmarks. Code and trained models are made available online at https://github.com/molecule-one/megan.