Computational prediction of enzyme mechanism and protein function requires accurate physics-based models and suitable sampling. I will describe how we have endeavored to balance these two requirements. Non-heme Fe(II) halogenases halogenate unactivated carbon atoms of diverse substrates at ambient conditions with exquisite selectivity seldom matched by non-biological catalysts. Using experimentally guided molecular dynamics (MD) simulations augmented with multi-scale (i.e., QM/MM) simulations, we have investigated and identified the substrate/active-site dynamics that enable selective halogenation. Limitations to both conventional density functional theory (DFT) and classical MM force fields remain for describing non-covalent interactions in comparison to experiment or wavefunction theory. I will describe how we have used a combined data mining and correlated wavefunction theory approach to understand the presence of privileged hydrogen bonding interactions in protein structures. Finally, I will discuss how we have begun to understand the effect of the dynamic nature of these interactions on the electronic structure of small proteins.
Last update: 10/15/2021, Ralf Bundschuh