THEORETICAL STUDIES OF PROTEIN FOLDING

该提案描述了蛋白质二级结构预测的新方法，三级折叠到低分辨率使用减少蛋白质模型，和结构细化使用原子水平势函数和溶剂化自由能的连续模型。 该方法将被应用于从头预测蛋白质结构，线程，环的同源性建模，并从稀疏的NMR数据集的蛋白质结构的确定细化。 具体的方法学领域包括：（1）二级结构预测的新评分函数;（2）三级折叠的约化势函数和全局优化方法的发展;（3）使用从头算量子化学数据构建可极化蛋白质势函数;（4）Poisson-Boltzmann和广义Born连续溶剂化模型的改进;（5）在三级折叠方法中加入NMR限制。

This proposal describes new methods for protein secondary structure prediction, tertiary folding to low resolution using reduced protein models, and structural refinement using atomic level potential functions and a continuum model for the solvation free energy. The methods will be applied to ab initio prediction of protein structure, threading, loop refinement in homology modeling, and protein structure determination from sparse NMR datasets. Specific methodological areas of focus include: (1) new scoring functions for secondary structure prediction; (2) development of a reduced potential function, and global optimization methods for tertiary folding; (3) construction of a polarizable protein potential function using ab initio quantum chemical data; (4) improvement of Poisson-Boltzmann and generalized Born continuum solvation models; (5) inclusion of NMR constraints in the tertiary folding methodology.