Structure Prediction of Helical Transmembrane Proteins

螺旋跨膜蛋白的结构预测

• 批准号: 7194520
• 负责人: YING XU
• 金额: $ 28.03万
• 依托单位: UNIVERSITY OF GEORGIA
• 依托单位国家: 美国
• 财政年份: 2007
• 资助国家: 美国
• 起止时间: 2007-02-01 至 2011-01-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/7194520
• 关键词: Adopted; Algorithms; Alzheimer' s Disease; Amyloid; Area; Aspartic Endopeptidases; Biological Models; Chemicals; Classification; Clinical; Code; Collection; Computational algorithm; Data; Databases; Development; Ensure; Etiology; Helix (Snails); Homologous Gene; Housing; Integral Membrane Protein; Investigation; Linux; Lipids; Medical; Membrane; Membrane Proteins; Methods; Modeling; Molecular Conformation; Nuclear Magnetic Resonance; Numbers; Pharmaceutical Preparations; Phase; Procedures; Process; Protein Precursors; Proteins; Protocols documentation; Purpose; Relaxation; Research; Research Personnel; Residual state; Running; Sampling; Shapes; Source; Staging; Structural Protein; Structure; System; Techniques; Testing; Vertebral column; Work; analog; base; beta secretase; beta-site APP cleaving enzyme 1; computerized tools; disorder control; ear helix; falls; improved; novel; nuclear Overhauser enhancement; programs; protein folding; protein structure prediction; research and development; research study; secretase; simulation; software development; statistics; tool

DESCRIPTION (provided by applicant): We propose to develop a suite of computational tools for the prediction of tertiary structures of alpha-helical transmembrane (TM) proteins. The core of the-project is to use experimental data-constrained Monte Carlo (MC) simulation to predict the 3D packing of TM helices. To maximally ensure that the MC simulation converges to the global minimum state of the energy suface, we propose to adopt the popular Wang-Landau (WL) algorithm in our implementaion of the MC simulation. The key feature of WL is the elimination of thermally activated energy barriers to possible conformation changes. Our initial folding prediction will be based on existing energy functions such as CHARMM. A new set of energy functions will be developed based on chemical/physical principles and known structures to facilitate more efficient folding calculation. To overcome the problem of searching the enormous conformational space for finding an optimally folded structure, novel MC sampling techniques specifically-tailored for TM proteins will be developed, one of which is to use stable hairpin structures formed by sequentially-neighboring helix pairs as building blocks for the assembly of helical bundles. Also, we propose to develop a systematic approach to generate a small number of distance, orientation, and geometric shape constraints through nuclear magnetic resonance (NMR) and other experiments such as paramagnetic relaxation enhancement, and to apply these experimental data to constrain the search space in the MC simulation. In addition, we propose to develop a threading-based prediction method for membrane proteins, which relies on solved structural homologues or analogs in the Protein Data Bank. Such predictions should generally provide good predictions of the topological arrangements of the TM helices, which could then be used as the starting structures of the MC simulations. The research proposed here, if successful, will provide a never before available capability for protein structural prediction to clinical researchers working on TM proteins, which are the targets for about 2/3 of the contemporary medical drugs. In addition, some of the proposed test systems, such as the amyloid precursor protein (APP) and the beta secretase protein (BACE1) are membrane proteins implicated in the etiology of Alzheimer's disease. The predicted structural information for these proteins could provide a basis for understanding and controlling this disease.

描述（由申请人提供）：我们建议开发一套计算工具，以预测α-螺旋跨膜（TM）蛋白的三级结构。项目的核心是使用实验数据约束的蒙特卡洛（MC）模拟来预测TM螺旋的3D填料。为了最大程度地确保MC模拟会融合到能源SUFAFE的全球最低状态，我们建议在我们对MC模拟的实施中采用流行的Wang-Landau（WL）算法。 WL的关键特征是消除了可能变化的热活化的能屏障。我们最初的折叠预测将基于现有的能量功能，例如Charmm。将根据化学/物理原理和已知结构开发一组新的能量功能，以促进更有效的折叠计算。为了克服搜索最佳折叠结构的巨大构象空间的问题，将开发出专门针对TM蛋白的新型MC采样技术，其中一个是使用稳定的发夹结构，由稳定的发夹结构由依次延迟的螺旋螺旋对作为螺旋大管组成的构建块形成。同样，我们建议开发一种系统的方法来通过核磁共振（NMR）以及其他实验（例如顺磁性松弛增强）产生少量距离，方向和几何形状约束，并应用这些实验数据来限制MC模拟中的搜索空间。此外，我们建议为膜蛋白开发一种基于螺纹的预测方法，该方法依赖于蛋白质数据库中解决的结构同源物或类似物。这样的预测通常应该对TM螺旋的拓扑布置进行良好的预测，然后可以将其用作MC模拟的起始结构。如果成功的话，这里提出的研究将为从事TM蛋白质的临床研究人员提供从未有的可用能力，这是当代医学药物的大约2/3的靶标。此外，一些提出的测试系统，例如淀粉样蛋白前体蛋白（APP）和β泌尿蛋白蛋白（BACE1）是膜蛋白，涉及阿尔茨海默氏病的病因。这些蛋白质的预测结构信息可以为理解和控制这种疾病提供基础。