Machine learning/molecular mechanics enzymology for the next generation of computational enzymatic catalysis

While quantum mechanics/molecular mechanics (QM/MM) frameworks have long enabled simulations of chemical reactivity, recent advances in machine learning interatomic potentials (MLIPs) have extended these capabilities by providing near-quantum accuracy at molecular mechanics efficiency. Embedded within machine learning/molecular mechanics (ML/MM) frameworks, MLIPs enable large-scale reactive simulations that were previously impractical with conventional QM/MM approaches. This perspective summarizes the datasets and training strategies used for reactive MLIPs, reviews recent progress in ML/MM-based enzymatic simulations, and discusses their potential extension to more complex scenarios, thereby identifying key opportunities and challenges for future research and applications.