111
Modeling of Biomolecular Systems Interactions, Dynamics, and Allostery Poster Session II
56-POS
Board 9
Dynamical Correlations and Long-Range Allostery Revealed by Intra-Protein Conditional
Activity
Milo Lin
.
University of California, Berkeley, Berkeley, CA, USA.
A new type of statistical measure, called the conditional activity, is introduced to quantify
correlations in the waiting times between transitions amongst the degrees of freedom. The
conditional activity is the dynamical analogue of the mutual information, and, for systems in
thermal equilibrium, correspond to the instantaneous change of the free energy barrier of one
degree of freedom following a transition in another degree of freedom. For three different
proteins, we calculated the conditional activity of side-chain fluctuations from 3 microsecond-
long all-atom molecular dynamics simulations. We show that the conditional activity reveals a
rich dynamical architecture, and that the principle eigenvector of this measure can retrodict
known pathways of dynamical allostery. The applicability of this method to predicting novel
allosteric pathways in proteins, as well as to other types of networks, will be discussed.
