Screened non-bonded interactions in native proteins manipulate optimal paths for robust residue communication

TL;DR

This study models protein residue networks with weighted edges based on amino-acid potentials, revealing how selective edge screening influences optimal communication paths and robustness of residue interactions.

Contribution

It introduces a novel edge screening method based on amino-acid potentials to analyze residue communication pathways in proteins.

Findings

Reduced networks with strongest interactions replicate full network shortest paths.

A constant rate of APL change persists until a critical link removal point.

Different perturbations lead to diverse optimal communication paths.

Abstract

A protein structure is represented as a network of residues whereby edges are determined by intra-molecular contacts. We introduce inhomogeneity into these networks by assigning each edge a weight that is determined by amino-acid pair potentials. Two methodologies are utilized to calculate the average path lengths (APLs) between pairs: To minimize (i) the maximum weight in the strong APL, and (ii) the total weight in the weak APL. We systematically screen edges that have higher than a cutoff potential and calculate the shortest APLs in these reduced networks, while keeping chain connectivity. Therefore, perturbations introduced at a selected region of the residue network propagate to remote regions only along the non-screened edges that retain their ability to disseminate the perturbation. The shortest APLs computed from the reduced homogeneous networks with only the strongest few non-bonded pairs closely reproduce the strong APLs from the weighted networks. The rate of change in the APL in the reduced residue network as compared to its randomly connected counterpart remains constant until a lower bound. Upon further link removal, this property shows an abrupt increase, towards a random coil behavior. Under different perturbation scenarios, diverse optimal paths emerge for robust residue communication.