Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces MAELLE, a novel approach to chemical reaction prediction that models reactions as discrete flow matching over electron occupation vectors, rather than relying on traditional methods like product generation or heuristic graph edits. By formulating the reactant-to-product mapping as a Continuous-time Markov Chain (CTMC) on a graph-structured integer-valued electron occupation space, MAELLE generates mechanistically interpretable edit trajectories without the need for elementary step annotations. The method demonstrates competitive performance on the USPTO-480K benchmark and exhibits robustness in out-of-distribution scenarios, effectively recovering mechanistic trajectories and predicting side products.
MAELLE reveals that modeling electron flow can yield more interpretable and robust predictions in chemical reactions than conventional graph-based methods.
Chemical reactions are fundamentally transformations in electron space, yet most machine learning approaches model them either through \textit{de novo} generation of product molecules or through heuristic graph edits that operate directly on molecular topology. We introduce MAELLE (\textbf{M}ech\textbf{A}nistic \textbf{E}dit f\textbf{L}ow-matching on e\textbf{L}ectron r\textbf{E}arrangements), which instead models reactions as discrete flow matching over electron occupation vectors. Concretely, we formulate the reactant-to-product mapping as a Continuous-time Markov Chain (CTMC) over the graph-structured integer-valued electron occupation space defined on all bonding, non-bonding, and hydrogen sites. To construct the intermediate edit trajectories, we generalize the discrete flow matching mixture path to discrete electron rearrangements using Optimal Transport, yielding a sequence of mechanistically interpretable edit moves without requiring elementary step annotations. MAELLE achieves competitive performance on the USPTO-480K benchmark compared with leading reaction prediction models. Beyond in-distribution accuracy, we evaluate robustness across two out-of-distribution settings - structural complexity and reaction type - and find that MAELLE maintains strong performance where existing methods degrade. Finally, because the learned flow operates over the full electron redistribution, MAELLE naturally recovers mechanistic trajectories that align with known chemistry and can predict side products of a reaction.