earticle

영어

t is a common belief that some residues of a protein are more important than others. In some cases, point mutations of some residues make butterfly effect on the protein structure and function, but in other cases they do not. In addition, the residues important for the protein function tend to be not only conserved but also coevolved with other interacting residues in a protein. Motivated by these observations, we propose that there is a network composed of the residues, the residue-residue interaction network (RRIN), where nodes are residues and links are set when the interaction strengths between residues are sufficiently large. We build the RRIN for the 40 diverse protein families. The interaction strengths are calculated by using McBASC algorithm. After constructing the RRIN, we identify residues that have high degree of connectivity (hub nodes), and residues that play a central role in network flow of information (information centrality nodes). We show that these residues are likely to be functionally important residues. Unlike other similar methods, our method is solely based on sequences. Therefore, the method can be applied to the function annotation of a wider range of proteins.