COMPUTER SIMULATIONS OF PROTEIN STRUCTURE AND DYNAMICS

这项研究的目标是发展更精确的溶剂化蛋白质的分子动力学模拟方法，并将其应用于生物物理化学中的问题，包括蛋白质在溶液中的静电效应的模拟，与蛋白质的配体结合的热力学，以及蛋白质的长时间动力学。有四个具体的目标：1.使用详细的原子模拟来表征蛋白质的介电性质：目前用于预测控制蛋白质活性和稳定性的静电性质的模型需要在定性和定量上进行改进。我们将开发一个一致的框架来模拟蛋白质的介电性质，可以弥合目前连续统和显式溶剂观点之间的差距。2.确定蛋白质-配体结合自由能的能量和熵分量自由能自由能模拟已被用于确定水中溶质的化学势和配体与蛋白质的结合自由能，但由于严格的计算机要求和理论障碍，还没有分析相应的热变和熵变化。结合热和结合熵为理解结合反应提供了重要信息。我们将把我们最近开发的从有机溶质在水中的自由能模拟中提取热焓和熵的方法扩展到蛋白质-配体结合反应。我们将探讨活性中心和配体柔性在结合热力学中的作用。3.建立了溶剂化蛋白质分子动力学模拟的隐式溶剂模型。隐式溶剂模型的计算速度使其在需要大范围的构象采样或较长的模拟时间的情况下非常适合用于溶剂化蛋白质的分子动力学模拟。我们将开发一个基于广义Born框架的隐式溶剂模型，该模型可用于蛋白质的分子动力学模拟。4.利用计算机模拟和核磁共振现象的模拟来研究蛋白质动力学。我们将继续与核磁共振小组在蛋白质动力学研究方面进行合作。这包括进一步开发更复杂的统计方法来从核磁共振弛豫实验中提取动力学参数，以及开发使用广义Born隐含溶剂模型和朗之万运动方程来模拟蛋白质长时间动力学的有效方法。

The goal of this research is the development of more accurate methods for molecular dynamics simulations of solvated proteins, and their application to problems in biophysical chemistry involving the modeling of protein electrostatic effects in solution, the thermodynamics of ligand binding to proteins, and the long time dynamics of proteins. There are four specific aims: 1. Characterize protein dielectric properties using detailed atomic simulations: Qualitative and quantitative improvements are needed in the models currently used to predict electrostatic properties that govern protein activity and stability. We will develop a consistent framework for modeling protein dielectric properties that can bridge the current gap between continuum and explicit solvent viewpoints. 2. Determine the energy and entropy components of protein-ligand binding free energies Free energy simulations have been applied to determine the chemical potentials of solutes in water and to determine the binding free energies of ligands to proteins, but the corresponding enthalpy and entropy changes have not been analyzed because of stringent computer requirements and theoretical hurdles. Binding enthalpies and entrophies provide important information to help understand binding reactions. We will extend methods we recently developed to extract enthalpies and entrophies from free energy simulations of organic solutes in water to protein-ligand binding reactions. We will explore the role of active site and ligand flexibility in the binding thermodynamics. 3. Develop implicit solvent model of molecular dynamics simulations of solvated proteins The computational speed of implicit solvent models makes them attractive for use in molecular dynamics simulations of solvated proteins when either extensive conformational sampling or long simulation times are required. We will develop an implicit solvent model based on the Generalized Born framework which can be used in molecular dynamics simulations of proteins. 4. Study protein dynamics using computer simulations and the modeling of NMR phenomena. We will continue our collaboration with NMR groups on the study of protein dynamics. This includes the further development of more sophisticated statistical methods to extract dynamical parameters from NMR relaxation experiments and the development of efficient methods for simulating the long time dynamics of proteins using the Generalized Born implicit solvent model and Langevin equations of motion.