GraSp-PSN: A web server for graph spectra based analysis of protein structure networks

Area of Science:

• Structural biology
• Bioinformatics
• Network science

Background:

• Protein function is intrinsically linked to conformational changes driven by ligand binding or environmental factors.
• Traditional protein structure analysis often focuses on atomic contacts, potentially overlooking emergent global structural properties.
• Understanding these global properties is crucial for a comprehensive analysis of protein behavior.

Purpose of the Study:

• To introduce GraSp-PSN, a web server for analyzing protein structure networks (PSN).
• To enable the acquisition of network parameters from atomic to global levels.
• To facilitate the comparison of protein structures and identification of critical residues through perturbation scoring.

Main Methods:

• Utilizing graph spectra analysis to represent protein structures as networks.
• Developing a web server (GraSp-PSN) to process 3D protein coordinates.
• Implementing algorithms for weighted PSN generation, structure comparison scoring, and residue perturbation analysis.

Main Results:

• GraSp-PSN provides weighted PSNs and network parameters from atomic to global scales.
• The server generates comparison scores reflecting local and global connectivity differences between protein structures.
• Residue perturbation scores are assigned, prioritizing key residues and their interactions within clusters.

Conclusions:

• GraSp-PSN offers a novel approach to protein structure analysis by incorporating network properties.
• The server's methods are versatile and applicable to network comparisons across various scientific disciplines.
• This tool enhances the understanding of protein structure-function relationships and facilitates comparative structural biology.