Allostery is an intrinsic dynamic phenomenon that underlies functional long-distance interactions in proteins, which we study here by stochastic calculus approach to elastic network models (ENMs). We show that once you drop the usually accepted high friction limit and include hydrodynamic interactions in ENMs, a simple measure that uses the pairwise difference in the time-delayed correlations of residue fluctuations provides insight about functional sites and their dynamical behaviour in allosteric communication. We present this with three exemplary cases Aspartate Carbamoyl transferase, Insulin Receptor and DNA-dependent Protein Kinase. We show that proteins possess characteristic pathways operating at different time-delay windows with slow to faster motions underlying the protein function. As these pathways help communication between key residues of functionality, they can also be used to identify their locations without any prior knowledge other than the protein crystal structure.

Dynamic Order in Allosteric Interactions