MULTI-SCALE PROTEIN STRUCTURE MODELING SIMULATION, AND PREDICTION

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. The proposed research involves four projects, described below, related to protein structure modeling, simulation, and prediction: (1) MULTI-RESOLUTION AND MULTI-LENGTH-SCALE SIMULATION OF SUPERMOLECULAR COMPLEXES: Large-scale conformational transitions in protein structures play an important role in a variety of cellular processes. Understanding such transitions is a central task of modern biophysics and structural biology. This project applies molecular dynamics simulation to model the motions of systems that involve coordinated large domain movements, such as the molecular chaperonin GroEL and F1-ATPase. (2) STRUCTURAL REFINEMENT FOR X-RAY CRYSTALLOGRAPHY USING NORMAL MODES: Traditional crystallographic refinement on X-ray data of proteins is limited by the enormous number of parameters required to describe the thermal fluctuations of the atoms. The use of normal mode vectors can significantly reduce the meaningful parameter set because the normal modes represent the intrinsic motions of the protein. This project aims to develop a software program that employs normal modes for crystallographic refinement. In addition to the benefits of reduced computational complexity, the normal mode model will provide more insight into protein dynamics because the anisotropic temperature factors calculated from the normal modes capture the motion inherent in the structure of the protein. (3) TOP-DOWN APPROACH TO PROTEIN STRUCTURE PREDICTION: Protein structure prediction is one of the most challenging areas in theoretical biophysics, and it plays an essential role in structural genomics and rational drug design. We recently developed OPUS, a conceptually new method that employs multi-scale, multi-layer and top-down prediction strategies. OPUS combines template-based and de novo methods to predict 3D protein structures from primary sequences. (4) PROTEIN DOMAIN RECOGNITION: Protein domain recognition is valuable for protein structure analysis, as it enables large proteins to be broken down into smaller, more palatable sub-structures that can be solved by experimental or computational techniques. Our ab initio domain recognition method has already demonstrated remarkable success in determining protein domain boundaries. This study aims to extend the prediction accuracy of the ab initio method by using a variety of experimental constraints, such as X-ray solution scattering data.