A Theoretical Study Of Human Influenza Neuraminidase (Subtype N1) And Its Inhibitors By Combining Homology Modeling, Molecular Docking And CoMFA 3D-QSAR Modeling

Recent the worldwide spread of A/H5N1 avian influenza with highly virulent has highlighted the potential threat of pandemic influenza in humans. Oseltamivir (Tamiflu) and zanamivir (Relenza) are currently the only two anti-influenza drugs targeting the neuraminidase (NA) enzyme of the influenza virus. Reports of the emergence of drug resistance further make the development of new potent anti-influenza inhibitors a priority. In the current thesis, the three-dimensional structure of N1 subtype of human influenza type A virus was built first by homology modeling and molecular dynamic simulations, using different crystal structure as the template. Subsequently three-dimensional quantitative structure-activity relationships (3D-QSAR) with comparative molecular field analysis (CoMFA) were performed on its inhibitors based upon docking active conformations. The thesis consists of four chapters as follows:In the first chapter, influenza virus, influenza neuraminidase and research development of influenza neuraminidase inhibitors were briefly introduced.For the studies described in the second chapter, basic theory and computational methods of homology modeling, molecular mechanics, molecular dynamics simulations, molecular docking and three-dimensional quantitative structure-activity relationships (3D-QSAR) were briefly introduced.In the third chapter, the three-dimensional structures of N1 subtype of human influenza type A virus were built first by homology modeling and molecular dynamic simulations, using the crystal structure of N1 subtype of avian influenza virus and the crystal structure of N9 subtype of avian influenza virus as the template, respectively. Subsequently three-dimensional quantitative structure-activity relationships (3D-QSAR) with comparative molecular field analysis (CoMFA) were performed on its inhibitors based upon docking active conformations then the good predictive models have been established, respectively. We found that model 1 which was built using the crystal structure of N1 subtype of avian influenza virus as the template was more reasonable and reliable than model 2 which was built using the crystal structure of N9 subtype of avian influenza virus as the template, and cross-validated coefficients q~2 of the 3D-QSAR model of model 1 through FlexX docking is higher than of the 3D-QSAR model of model 2 through FlexX docking, then the predictive ability of 3D-QSAR model of model 1 will be more powerful, and it could be further employed as a complementary scoring-function in the coming structure-based (both receptor and ligand) virtual screening out novel potent virus inhibitors.In the last chapter, the summary and outlook of the full thesis were performed.