Modeling RNA Tertiary Structure Folding by a Hierarchical Framework

描述（由申请人提供）：RNA分子是参与许多基本生物学作用的重要细胞成分。由于RNA的结构特征与其生物学功能密切相关，因此从序列预测RNA结构具有很大的兴趣。提出的研究涉及RNA结构预测-使用图论和统计工具开发分层建模方法，以及RNA设计-工程荧光核糖开关，可以“感知”转录终止，为体内过程的研究提供可视化。提出的方法需要开发新的数学和统计/计算机科学工具，以便在3D中有效地采样RNA图对象-提供初始的，粗略的采样水平-并开发系统的基于结构的统计方法，使用数据挖掘工具对RNA构象进行评分。基于已知结构的RNA连接基序的广泛分析将用于开发单独的局部和全局统计势，以预测局部几何方向以及全局远程相互作用。这些RNA模型和相关电位将在以下分层设计的三个相互关联的折叠阶段进行组合和仔细测试：(1)使用MC采样在树图上粗略地探索RNA的结构构象空间，树图表示嵌入在具有选择RNA样构象的统计电位的3D晶格中的RNA二维结构；(2)利用具有更高统计潜力的粗粒度每核苷酸3头模型来指导3D组装，提高预测精度；(3)利用具有力场势的全原子模型的动力学模拟来细化最佳候选模型。经过仔细的分析，开发和测试新的数学工具，计划的每个阶段将被细化，最终集成组件。通过将荧光适体的已知元素与控制基因表达的核糖开关的基于配体的构型重排相结合，将追求荧光核糖开关的新设计应用。

DESCRIPTION (provided by applicant): RNA molecules are important cellular components involved in many fundamental biological roles. Because the structural features of RNA are intimately connected to their biological function, there is great interest in predicting RNA structure from sequence. The proposed research addresses both RNA structure prediction - development of a hierarchical modeling approach using graph theory and statistical tools, and RNA design - engineering fluorescent riboswitches that can "sense" transcription termination, offering researches visualization of the process in vivo. The proposed approach requires development of new mathematical and statistical/computer science tools for sampling RNA graph objects efficiently in 3D - to provide an initial, coarse level of sampling - and for developing systematically structure-based statistical approaches to score RNA conformations using data-mining tools. Extensive analysis of RNA junction motifs based on known structures will be used to develop separate local and global statistical potentials to predict both local geometric orientations, as well as global long-range interactions. These RNA models and associated potentials will be combined and carefully tested in three inter-related stages of folding in the following hierarchical design: (1) Explore RNA's structural conformation space coarsely using MC sampling on tree graphs that represent RNA 2D structures embedded in 3D lattices with statistical potentials that select RNA-like conformations; (2) Improve prediction accuracy using a coarse-grained three-bead-per-nucleotide model with a higher-level statistical potential to guide 3D assembly; and (3) Refine the best candidates using dynamics simulations on full atomistic models with force-field potentials. After careful analysis followed by development and testing of the new mathematical tools, each stage of the plan will be refined to finally integrate the components. New design applications for fluorescent riboswitches will be pursued, by marrying known elements of fluorescent aptamers with ligand-based configurational rearrangements of riboswitches that control gene expression.