Coarse-Grained Models of Proteins

描述（由申请人提供）：蛋白质运动的许多方面可以用粗粒度模型来理解。我们的假设是，解释蛋白质行为的许多方面不需要原子细节，这种简化可以促进更深层次的理解。总体目标是了解蛋白质的运动和功能是如何由结构控制的，为什么蛋白质序列折叠成一组有限的结构，并建立紧密包装的作用和蛋白质的形状在它们的运动。在这个项目中，我们将研究运动、形状、结构、相互作用和协作水平之间的关系。目的1：用弹性网络建模蛋白质动力学。我们将使用弹性网络模型来研究蛋白质如何将它们的运动限制到对功能最重要的运动。正态分析将以高计算效率来识别这些重要的功能运动，以发展分子机制。我们将研究酶活性位点的原子运动，以了解大结构域运动如何控制原子运动。我们将使用弹性网络来解释单分子拉扯实验，并预测蛋白质解开的顺序。初步结果表明，弹性网络模型不仅适用于固有构象周围的波动，而且适用于施加外力使蛋白质变形并破坏其固有接触时产生的瞬态。这些结果表明，结构控制着蛋白质的整体运动，即使是短暂的状态。为了进一步验证这一假设，我们将进行更多的单分子拉模拟，以及折叠路径上瞬时蛋白质构象的结构分析。弹性模型取得的主要成功依赖于填料密度和蛋白质形状的良好表示，我们将在Aim II中进行研究。目的二：建模蛋白质包装和相互作用的协同性。蛋白质中残基的密集堆积是其最重要的特征之一。我们计划继续研究内部包装。新电位的重点将放在蛋白质中氨基酸的相对取向上。我们将开发多体接触电位，以识别穿线中诱饵之间的天然结构，并研究蛋白质中邻近残基簇内的取向分布，使用规则多面体（如二十面体）或加泰罗尼亚固体（如四六面体）。我们的基本原理是使用各种多面体模型来理解蛋白质包装和氨基酸相互作用，以开发改进的多体电位。更好地理解蛋白质内部相互作用的协同性是非常重要的，因为这直接影响到蛋白质运动和对力的反应方式。这两个目标是高度相互关联的，将显著推进我们对蛋白质结构、动力学和功能的认识。

DESCRIPTION (provided by applicant): Many aspects of protein motion can be comprehended with coarse-grained models. Our hypothesis is that atomic detail is not required to explain many aspects of protein behavior, and this simplification can facilitate a deeper understanding. The overall goal is to develop an understanding of how protein motions and function are controlled by structure, why protein sequences fold to a limited set of structures, and to establish the roles of tight packing and the shapes of proteins on their motions. In this project we will investigate the relationships among motions, shapes, structures, interactions and levels of cooperativity. Aim I: Modeling protein dynamics with Elastic Networks. We will use elastic network models to study how proteins restrict their motions to the motions most essential for function. Normal mode analyses will be performed to discern these important functional motions with high computational efficiency to develop molecular mechanisms. We will investigate the atomic motions in active sites of enzymes to see how the large domain motions control the atom movements. We will use elastic networks to interpret single molecule pulling experiments and predict the order in which proteins unravel. Preliminary results show that elastic network models are applicable not only to fluctuations around native conformations, but also to transient states arising when an external force is applied to deform a protein and break its native contacts. These results suggest that structure controls the global motions of proteins, even for transient states. To further verify this hypothesis we will perform more single molecule pulling simulations, and structural analyses of transient protein conformations along folding pathways. The major successes achieved with the elastic models rely upon having good representations of the packing density and protein shape, which we will investigate in Aim II. Aim II: Modeling Protein Packing and Cooperativity of Interactions. Dense packing of residues in proteins is one of their most important characteristic features. We plan to continue our studies of internal packing. The emphasis for new potentials will be on the relative orientations of amino acids in proteins. We will develop many-body contact potentials for identifying native structures among decoys in threading, and also study orientational distributions within clusters of nearby residues in proteins, using regular polyhedra such as icosahedra, or Catalan solids such as tetrakis hexahedra. Our rationale is to use various polyhedral models to comprehend protein packing and amino acid interactions for developing improved many-body potentials. A better understanding of the cooperativity of interactions within proteins is extremely important because this directly influences the ways in which proteins move and respond to forces. Both Aims are highly interconnected and will significantly advance our knowledge of protein structure, dynamics and function. PUBLIC HEALTH RELEVANCE: Success in this project will affect many fields of molecular science - from the selection of protein targets for drug design to a general comprehension of how cells function. Progress on this project is critical for developing ways to meaningfully simulate cellular components and to utilize the rapidly growing cell imaging data. Improving the abilities to model protein motions can impact public health in important ways by enhancing our basic understanding of protein behavior and by facilitating better, more effective selection of protein targets for drug design.