Prediction of biomolecule kinetics using physics-based Brownian dynamics to data-driven machine learning methods

We present a comprehensive review of biomolecular binding kinetics and their modeling via Brownian dynamics simulations, with particular emphasis on enzyme-substrate interactions in cellular environments. We outline the theoretical foundations of Brownian dynamics (BD) and its application to modeling association and dissociation processes in both homogeneous and heterogeneous media. We further examine the role of BD in relation to emerging machine learning (ML) approaches for directly predicting binding kinetics. Finally, we propose that BD simulations can serve as a critical bridge between molecular-scale models and continuum, cell-level descriptions, offering a pathway toward a multiscale understanding of in vivo kinetic phenomena.