Ab Initio Structure Prediction of Polypeptides

描述(申请人提供)：本项目的总体目标是提高我们对单链多肽从头算结构预测的理解。基于我们在以下方面的最新进展：(1)溶剂化寡肽的自由能计算(Klepeis和Floudas，1999)；(B)通过AlphaBB方法的确定性全局优化(Adjiman等人，1998a，b；Floudas，2000)及其对溶剂化寡肽的应用(Klepeis等人，1998；Klepeis和Floudas，1999b)；以及(C)通过稀疏核磁共振限制条件对寡肽的结构进行精细化(Klepeis等人，1999)，拟议的研究旨在不使用数据库中可获得的信息而从氨基酸序列中确定多肽的二级和三级结构。我们计划重点实现以下目标： (A)研究一种识别单体多肽中螺旋链段的新方法，该方法基于重叠五肽的熵和自由能计算，并通过非线性Poisson-Boltzmann方程处理静电学。 (B)研究一种基于混合整数优化框架的新方法，用于确定反平行、平行和多标准的β-折叠，以及识别二硫键。 (C)研究一种基于确定性全局最优化和扭角动力学的新方法，用于在存在稀疏限制的情况下精制多肽的结构，这些限制要么来自核磁共振实验，要么通过由(A)和(B)中提出的方法提供的二级结构信息。 (D)基于(A)、(B)和(C)中的进展，研究单链多聚海潮从头算结构预测的新框架。 (E)开发(A)、(B)、(C)和(D)的分布式计算算法，将其应用于多肽，并开发用于多肽三维结构从头算预测的工具。

DESCRIPTION (provided by applicant): The overall aim of this project is to improve our understanding of the ab initio structure prediction of single chain polypeptides. Based upon our recent advances in (1) free energy calculations of solvated oligopeptines (Klepeis and Floudas, 1999); (b) deterministic global optimization through the alphaBB approach (Adjiman et al., 1998a,b; Floudas, 2000) and its application to solvated oligopeptides (Klepeis et aL, 1998; Klepeis and Floudas, 1999b); and (c) structure refinement of oligepeptides through sparse NMR restraints (Klepeis et al., 1999), the proposed research is directed at determining the secondary and tertiary structure of polypeptides from the sequence of amino acids without using information available from the databases. We plan to focus on the following objectives: (a) Study a new approach for the identification of the helical segments in monomeric polypeptides that is based on entropic and free energy calculations of overlapping pentapeptides, and treatment of the electrostatics through the nonlinear Poisson-Boltzmann equation. (b) Investigate a novel approach, based upon a mixed-integer optimization framework, for the determination of antiparallel, parallel and multistanded beta sheets, as well as the identification of the disulfide bridges. (c) Study a new approach, based on deterministic global optimization and torsion angle dynamics, for the structure refinement of polypeptides in the presence of sparse restraints that arise either from NmR experiments or through secondary structure information provided by approaches such as those proposed in (a) and (b). (d) Investigate a novel framework for the ab initio structure prediction of single chain polypep tides based on the advances in (a), (b) and (c). (e) Develop distributed computing algorithms for (a), (b), (c) and (d), apply them to polypeptides, and develop tools for the ab initio prediction of three-dimensional structures of polypeptides.