BPS2026 – Coarse-grained modeling of protein-DNA interactions: Reproducing experimental binding affinity

Understanding protein-DNA interactions is fundamental to elucidating gene regulation, restriction-modification systems, and genome organization. Restriction enzymes, as a model class of DNA-binding proteins, provide a controlled system to study such interactions. In this work, we develop coarse-grained models of protein-DNA complexes using the ARBD framework (atomic resolution Brownian dynamics) and the shape-based coarse-graining and mrDNA models to represent protein and DNA. The protein-DNA binding interactions are parameterized iteratively against experimentally measured dissociation constants (Kd), allowing us to simulate association and dissociation processes. For complexes that undergo conformational changes upon binding—such as for proteins that wrap around or bend the DNA—we show that fine-tuning the protein bond network against atomistic distributions of the apo protein enables us to capture both binding affinity and conformational dynamics. Together, these advances provide a robust framework for studying DNA-protein interactions at the mesoscale and extend the applicability of coarse-grained models to shape-dependent protein-DNA interactions.