@ppxasjsm.bsky.social and I were delighted to collaborate on this with Marko Hanževački, Papu Kalita and @adrianmulholla1.bsky.social

… an upgraded Amber force field for zinc metalloproteins taken from literature. The correlation is significantly improved, but the average errors are greater when compared to the restraint approach. Check out the full paper for analysis of outliers and extensive QM/MM and MM/MD simulations!

For VIM-2, a metallo-β-lactamase, the first model we evaluated was a distance restraint -based model. While the results are somewhat predictive, there is still room for improvement, which we already saw in the second model evaluated which is…

For KPC-2, a serine-β-lactamase, standard RBFE protocols and an Amber force field lead to excellent correlation with experiment and low mean errors.