Physics-Based Protein\nNetworks Might Recover Effectful\nMutationsa Case Study on Cathepsin G

Molecular dynamics simulations have been remarkably effective\nfor\nobserving and analyzing structures and dynamics of proteins, with\nlonger trajectories being computed every day. Still, often, relevant\ntime scales are not observed. Adequately analyzing the generated trajectories\ncan highlight the interesting areas within a protein such as mutation\nsites or allosteric hotspots, which might foreshadow dynamics untouched\nby the simulations. We employ a physics-based protein network and\npropose that such a network can adequately analyze the protein dynamics.\nThe analysis is conducted on simulations of cathepsin G and neutrophil\nelastase, which are remarkably similar but with different specificities.\nHowever, a single mutation in cathepsin G recovers the specificity\nof neutrophil elastase. The physics-based network built on the interactions\nbetween residues instead of the distances can pinpoint the active\ntriad in the proteins studied. Overall, the network seems to capture\nthe structural behavior better than purely distance-based networks.