An Approach of Mutual Information-based Characterization of Allosteric Macromolecules

Allostery is a fundamental mechanism by which many biological conformational processes are controlled. In this paper, we view the activities of allosteric macromolecules as the dynamics of a spatial network, where the nodes represent intramolecular sites or residues, and the links are the physical connections or bonds. We discuss how mutual information, an information-theoretic quantity, could be used to characterize the spatial relationships among the different intramolecular sites caused by allosteric activities. We envision that such an approach could reduce the apparent complexity of the network by identifying the critical nodes and their interrelationships in allostery. Such a mapping from spatial network to information network may allow information bit representation of allosteric interactions, which coupled with machine learning tools may be used to study complex biological processes and develop new therapeutics.