Development of Chemical Bond based Elastic Network Model and its application in identifying functional motions in H5N1 highly pathogenic avian influenza viruses

Here we proposed a Chemical Bond based Elastic Network Model (CB-ENM) method, which was demonstrated with a better performance in representing functional movements, especially the local motions, than the conventional normal mode analysis. CB-ENM was applied in characterising functional motions in surface glycoprotein Haemagglutinin (HA) of H5N1 highly pathogenic avian influenza virus. Eleven local peaks were identified amongst HA fluctuation spectrum, and eight of these peaks were correlated with the reported five antigenic binding sites and Receptor Binding Sites (RBS) in influenza viruses. Our study showed that CB-ENM has a potential in identifying functional residues in protein structure.