论文部分内容阅读
本文基于图论(graph theory)思想挖掘功能相似蛋白质。将蛋白质相互作用网络(protein-protein interaction network,PPI)转化为图,使用1-hop方法把相互作用网络拆分为子图。应用最短路径法,依次把拆分得到的每一个子图与所有子图进行两两比对,寻找最优比对结果,挖掘物种内有相似功能的蛋白质。应用该方法对构建的人类蛋白质相互作用网络进行比对,共挖掘到1081对功能相似蛋白质。该方法的特点是数据量偏小、较可靠。计算结果中还出现了一些序列相似度较低、用序列比对的方法难以挖掘到的功能相似蛋白质。通过对结果的分析可以发现,趋同进化和同源有时很难区分。本实验的全部数据、程序和结果可在网站http://bina.biositemap.com上查询。
In this paper, functionally similar proteins are searched based on the graph theory. The protein-protein interaction network (PPI) was transformed into a map, using the 1-hop method to split the interaction network into subgraphs. Using the shortest path method, each subgraph obtained by the splitting is compared with all the subgraphs one by one to find the optimal alignment result and to excavate proteins with similar functions in the species. Using this method, the constructed human protein interaction networks were compared and 1081 pairs of functionally similar proteins were found. The method is characterized by a small amount of data, more reliable. In the calculation results, some functionally similar proteins that are less similar in sequence similarity and difficult to find by the sequence alignment method are also appeared. Analysis of the results shows that it is sometimes difficult to distinguish between evolutionary convergence and homology. All the data, procedures and results of this experiment can be found on the website http://bina.biositemap.com.