Centrality based analysis of amino acids network

Abstract

A network is a crucial asset in biology for capturing and exploring interaction data in biological systems of many types, such as protein-protein communications, amino acid associations, gene regulation, and cellular metabolism. In this article, we constructed an amino acid distance matrix by considering each base's positional relevance in a codon, chemical types: Purine and Pyrimidine, and H-bonding count. Based on the amino acid distance matrix, we eventually generated a twenty amino acid network having evolutionary significance. We reviewed multiple centrality metrics to assess the relative importance of amino acids in the proposed network: Degree Centrality, Closeness Centrality, Betweenness Centrality, Eigenvector Centrality, Eccentricity Centrality, and Radiality Centrality. We also looked at the correlation coefficients between the different centrality measures to figure out whether the network is assortative or disassortative. Furthermore, we examined the Clustering Coefficient and Degree Distribution as two effective network measures, and the results seem noteworthy.

Keywords

• amino acid
• genetic code
• codon redundancy
• centrality measure
• correlation
• clustering coefficient