New Algorithms for Detecting Sequence and Structure Simi

该项目分为3个子项目： 1.改进多序列比对的迭代自动方法。目前构建CDD数据库的起点是从SMART和PFAM等多个序列导入的蛋白质家族模型。由于所有已知的多序列比对算法都不是很完美，导入的比对存在误差，现在可以手动删除这些误差。该项目的目标是提供、减少和简化这项手工工作。 2.检测结构相似性的新方法。这一分项目的目标是开发新的方法，以便快速发现结构相似之处，以补充目前使用的方法(VASS)。因此，特别关注循环结构和索引方法的发展。 3.图论域间相似关系的应用。蛋白质序列或结构之间的相似关系可以方便地用图来表示。这个子项目的目标是研究这类图形的属性，并开发分析工具。

The project is divided into 3 subprojects: 1.Iterative automatic methods for refining multiple sequence alignment. Currently starting point of constructing CDD database is a multiple alignment imported form protein family models as SMART and PFAM. Since all known multiple sequence alignment algorithms are less than perfect the imported alignment have errors that are now removed by hand. The goal of the project is to provide reduce and simplify this manual work. 2.New approaches for detecting structural similarities. The goal of this subproject is to develop new approaches for fast detection of structural similarities that would complement currently used method (VAST). Thus particular attention is given to loop structure and developing of indexing methods. 3.Applications of graph theory similarity relations between domains. The similarity relation between proteins sequences or structures can be conveniently represented as a graph. The goal of this subproject is to study the properties of such graphs and develop tools for their analysis.