Computer Simulation Theory of Globular Protein Dynamics

描述(由申请人提供)：该项目解决了广泛的相互关联的生物学问题，利用基于结构的方法来预测生化功能：蛋白质结构和功能预测方法的拟议改进将导致更好的方法来选择药物先导和确定导致药物副作用的目标蛋白质相互作用。同样，由于DNA-蛋白质相互作用对包括转录在内的各种生物过程至关重要，因此将开发更好地预测DNA结合蛋白在与DNA的复合体中的身份和结构的方法。为了阐明亚细胞过程的定性行为，将进行示意性分子模拟。一个关键的想法是观察到，在进化上遥远的蛋白质中，有一些保守的结构和功能特征可用于配体和蛋白质的排序和精制；值得注意的是，即使在蛋白质侧链和配体功能基团(例如羟基)的水平上，这种保守也是成立的。为了实现这些目标，提出了五个具体目标：1.对Tasser进行改进和扩展，使其能够在原子细节上处理模型。2.利用多结构域全息蛋白的结构保守性，改进多结构域蛋白质结构预测。3.将改进蛋白质组中脱靶蛋白的配基排序/鉴定方法。4.对蛋白质结合DNA的DNA序列依赖性和四级结构的预测将得到改进。5.为了探索流体动力相互作用(被证明对细胞内分子扩散很重要)和拥挤在细胞内动力学中的作用，我们将对亚细胞过程进行理想化的布朗动力学模拟，如代谢途径、肌动蛋白组装机制、肌球蛋白沿肌动蛋白细丝运动的机制以及与转录相关的蛋白质动力学方面。在每个阶段，都会进行仔细和全面的基准测试；通常，这之后会将开发的方法学应用于蛋白质组。所有工具和数据库都将作为网络服务提供；所有软件都可以下载到本地服务器上使用。每个目标都是更全面目标的一部分，目的是增加我们对亚细胞过程和生化功能的了解，这将有助于加快药物发现，并提供一般的生化见解。

DESCRIPTION (provided by applicant): This project addresses a broad spectrum of interrelated biological problems that exploit structure based approaches to biochemical function prediction: Proposed improvements in protein structure and function prediction methodologies will lead to better approaches for drug lead selection and identification of of-target protein interactions responsible for drug side effects. Similarly, since DNA-protein interactions are crucial to a variety of biological processes including transcription, methods that better predict the identity and structure of DNA binding proteins in complex with DNA will be developed. Towards elucidation of the qualitative behavior of subcellular processes, schematic molecular simulations will be undertaken. A key idea is the observation that across evolutionarily distant proteins, there are conserved structural and functional features that can be used for ligand and protein ranking and refinement; remarkably, this conservation holds even at the level of protein side chain and ligand functional groups (e.g. hydroxyls). To realize these objectives, five Specific Aims are proposed: 1. TASSER will be improved and extended to treat models at atomic detail. 2. Exploiting structural conservation across multiple-domain holo-proteins, multidomain protein structure prediction will be improved. 3. Improved approaches for ligand ranking/identification of off-target proteins in proteomes will be developed. 4. Prediction of the DNA sequence dependence and quaternary structure of protein-bound DNA will be improved. 5. To explore the roles of hydrodynamic interactions (shown to be important for intracellular molecular diffusion) and crowding in intracellular dynamics, idealized Brownian Dynamics simulations of subcellular processes such as metabolic pathways, the mechanism of actin assembly, the mechanism of myosin motion along actin filaments, and aspects of protein dynamics associated with transcription will be undertaken. At every stage, careful and comprehensive benchmarking will be done; often, this will be followed by application of the developed methodology to proteomes. All tools and databases will be provided as web services; all software will be downloadable for use on local servers. Each Aim is part of a more comprehensive objective to increase our understanding of subcellular processes and biochemical function that will help accelerate drug discovery as well as provide general biochemical insights.