LSP-based Protein Residue Networks

Our method is based on the graph-theoretic representation of protein structure, commonly known as Protein Residue Networks (PRNs). While defining nodes of these networks as residues is straightforward, the definition of links between them varies significantly.

This figure shows a simple network with nodes connected by lines of different thickness. In traditional methods, thicker lines mean that two residues interact strongly, often based on cross-correlation or mutual information from MD simulations. In contrast, strong links in our PRNs reflect local stability of the residues. This directly addresses the local entropic properties of the protein. We, thus, shift the focus from interaction to stability, offering a new angle to analysis of protein entropy. Such definition of weights in PRNs allow us to build 'Thermal maps' of a protein that visualize distribution of protein entropy related to its thermal vibrations.

The figure here displays three states of the Catabolite Activator Protein (CAP): the apo state at the top, a state with a single cAMP molecule bound in the middle, and a state with two cAMP molecules bound at the bottom. Nodes that represent residues of low entropy are depicted with a darker color and larger diameter, indicating densely packed areas of stability.

CAP is a classic example of entropy-driven allostery, where the protein's function changes with no significant difference in protein structure. Our 'thermal maps' provide a direct and intuitively simple way to visualize entropic shifts that occur with consecutive cAMP binding, allowing us to see entropy-driven allostery in action.

Similar maps can be generated for any protein, offering valuable insights into global changes in protein entropy in response to ligand binding, mutations, post-translational modifications, or other computationally simulated changes. Importantly, the entropy analysis is not limited to amino acid residues; it can also include ligands (like cAMP, as shown in the figure), nucleotides, or lipids.