蛋白质相互作用网络的相似子网搜索问题研究

Study on similar subnet searching method of protein-protein interactions network

蛋白质相互作用网络(Protein-Protein Interactions Network,PIN)的相似性问题是目前生物信息学领域研究的热点。将计算机科学和生物学相结合,提出了蛋白质相互作用网络邻居优先搜索算法。该算法综合蛋白质的序列信息和蛋白质相互作用网络的拓扑结构信息,适度提高与相似蛋白质有直接相互作用的蛋白质之间的相似系数,实现了不同物种间蛋白质相互作用相似子网络的搜索。与同类算法的对比实验表明,该算法可以处理更大规模的目标子网搜索,计算速度明显提高,且利用该算法获得的结果与目标子网具有更长的相似路径。论文采用该算法研究了酵母和果蝇的蛋白质相互作用网络,获得了10条相对保守的蛋白质相互作用(Protein-Protein Interactions,PPI)。

The problem of the Protein-Protein Interactions Network（PIN） similar is hot in bioinformaties researching domain.Here, computer science and biology is combined to present a prior neighbor searching method of protein-protein interactions network. The method uses both the protein sequence and the PIN structure.In the method,similarity is added to the proteins that have interaction with similar proteins and similar subnet searching of deferent species protein-protein interactions is implemented.Comparing with other methods,the method can search larger object subnet in quicker speed and it can get a more similar and longer PPI.In this paper,the method is applied to the PIN of yeasts and flies and gets ten relative conservative PPIs.