Protein Fold Modeling and Recognition From Multiple Structures

本课题的目标是开发基于多结构信息的蛋白质折叠建模和蛋白质折叠识别方法。使用三维结构对齐方法，共享同一褶皱的多个结构将相互对齐。利用排列结构和结构训练数据集构建褶皱的统计模型。将设计方法来评估查询序列与折叠模型的兼容性。预测的查询序列的结构属性将被纳入评分方案的设计中。有效地利用多结构信息，可以提高折叠识别的成功率和叠线序列的准确性。如果成功，本项目的结果将有助于从大规模的DNA测序项目中阐明许多编码序列的功能和结构。

The objective of this project is to develop methods for protein fold modeling and protein fold recognition based on multiple structure information. Multiple structures sharing the same fold will be aligned to each other using 3-D structure alignment methods. A statistical model for the fold will be constructed using the aligned structures and a training data set of structures. Methods will be designed to assess the compatibility of a query sequence to the fold model. Predicted structural attributes of the query sequence will be incorporated into the design of scoring schemes. Effective utilization of multiple structure information is expected to increase the success rate of fold recognition and the accuracy in threading sequence to fold. If successful, the results of this project will help to elucidate the functions and structures of many coding sequences from the large scale DNA sequencing projects.