FUNCTION DIVERGENCE IN PROTEIN FAMILIES

这项研究的目的是研究功能相似性 和具有广泛相似结构的蛋白质之间的差异， 没有可识别的序列相似性。自动结构比较 程序，连同新一代的折叠预测算法， 为我们提供了许多蛋白质的例子， 它们属于同一个结构类。 与此同时， 结构相似性可以或可以不连接到任何功能 相似性 将建立蛋白质结构和功能的数据库， 允许搜索具有某些结构和 功能特点 然后将对这些群体进行分析， 序列中结构和功能重要的区域， 将它们与结构和功能相似性水平相匹配/ 发散 比较建模和结构的结合 蛋白质结构的分析将被用于并锻造成一个 一套标准化的理论工具和程序， 预测使用线程和序列分析，但其序列 差异大到足以预期显著的功能差异。 本项目所述研究的总体目标是 理解序列/结构/功能关系，或 选择蛋白质的水平，其中详细的预测功能 并且可以实现新测序的蛋白质的作用模式。 随着大规模基因组测序的不断涌现， 排序项目，它是确定这些可能的功能， 这些蛋白质使这些数据变得如此重要。

The goal of the proposed research is to study functional similarities and differences between proteins with broadly similar structures, but no recognizable sequence similarity. Automated structure comparison programs, together with a new generation of fold prediction algorithms, provide us with many examples of proteins that do, or are likely to, belong to the same structural class. At the same time, such broad structural similarity may or may not be connected to any functional similarity. A database of protein structure and function will be built, allowing a search for groups of proteins with certain structural and functional features. Such groups will then be analyzed to identify structurally and functionally important regions in the sequence and to match them with levels of structural and functional similarity/ divergence. A combination of comparative modeling and structure analysis of protein structures will be used and forged into a standardized set of theoretical tools and procedures that could be predicted using threading and sequence analysis, but whose sequence divergence is large enough to expect significant functional differences. The overall goal of the research described in this project is to bring an understanding of the sequence/structure/function relationship or selected proteins to a level where the detailed predictions of function and mode of action for newly sequenced proteins could be accomplished. With the flood of new sequences coming from large scale genome sequencing projects, it is identifying the possible function of these proteins that would make these data so important.