Five Hierarchical Statistical Models for Protein Folding

伦斯勒理工学院获得了研究蛋白质折叠途径的资助。利用数据挖掘和建模实验，该项目将开发5个统计模型来表示折叠蛋白质路径的不同阶段。模型表示序列-结构空间中的一棵决策树，每个模型包括以下步骤：从启动、折叠短基序；传播、结构化区域的局部扩展；凝聚、局部单元对之间的接触；填充、单元折叠成空间填充方式；到拓扑、环长排序等，以产生完全折叠的结构。绿色荧光蛋白将作为测试案例，策略性的点突变为中间结构的实验验证和方法的整体成功提供了一种手段。该项目将整合到RPI的本科教育中，并将为学生提供在实验室工作的鼓舞人心的体验。

Rensselaer Polytechnic Institute is awarded a grant to study the folding pathways of proteins. Using data mining and modeling experiments the project will develop 5 statistical models to represent the various stages of the path to a folded protein. The models represent a decision tree in the sequence-structure space, where each model feeds into the following step: from Initiation, folding of short motifs; Propagation, local extensions of structured regions; Condensation, contacts between pairs of local units; Packing, units fold into a space filling way; to Topology, ordering of loop lengths, etc. to yield the fully folded structure. Green Fluorescent Protein will be used as a test case, with strategic point mutations providing a means for experimental verification of intermediate structures and the overall success of the method. The project will be integrated in the undergraduate education at RPI and will offer stimulating experiences for the students to work in a lab.