A STUDY OF CORRELATED PROTEIN MOTIONS BY NORMAL MODE ANALYSES AND MOLECULAR DYN

点击翻译按钮获取中文摘要

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. Correlated protein motions are of great functional significance and have been the target of various experimental and computational studies. Both molecular dynamics (MD) simulation and normal mode analysis (NMA) are able to provide information about the internal protein dynamics. Starting from a limited sampling of the protein conformation space, usually a section of MD simulation trajectory, principal component analysis (PCA, or essential dynamics) teases out the correlation motions through a diagonalization of the covariance matrix for the relevant atom group. The direction and amplitude of the vibrational modes are represented by the resulting eigenvectors and the corresponding eigenvalues, respectively. On the other hand, based on a harmonic assumption of the protein energy surface, normal mode analysis (NMA) yields the direction and amplitude of the correlated motion through diagonalization of the Hessian matrix, which contains the second derivative of the protein energy surface. Due to a limitation of computational resources, NMA approaches at different resolutions, from simplest elastic network model (ENM), block normal mode (BNM), to the classical all-atom normal mode methods, have been used more often than the time consuming MD-based PCA approach. However, the harmonic assumption of the energy surface and the implicit treatment of solvent make the NMA difficult to reproduce the often diffusive and anharmonic nature of protein dynamics. Here I propose to carry out parallel NMA and MD simulations on selected high resolution X-ray structures and comparing the results from both computational approaches to the experimentally determined parameters, including the isotropic vibrational amplitudes and anisotropic vibrational directionality. It is expected that MD-based PCA would bring a closer match to the experimental observations than does NMA.