中文摘要：

关键蛋白质的识别有助于从系统水平上理解生命活动过程,基于蛋白质相互作用网络拓扑特征的关键蛋白质识别可以有效地提高识别精度和速度。通过蛋白质节点的最短路径数和点介数可以作为衡量其节点中心度的方法,但计算速度和计算规模有限。根据所预测蛋白质相互作用网络的特点,提出了基于最短路径技术的关键蛋白质识别方法,选择合理的识别阈值和拓扑参数,对全蛋白质相互作用网络的关键蛋白质进行预测。实验表明,所提出的识别方法可以有效描述蛋白质节点的重要性,在不影响计算精度的前提下,可对连通性好,边密度大的全蛋白质相互作用网络进行关键蛋白质识别。

英文摘要：

Identification of essential protein contributes to understanding the living processes at the system level.The essential protein identification based on protein interaction network topology characteristics can improve the accuracy and speed of recognition effectively.The number of shortest paths which pass a protein node and the vertex betweenness can be used to measure the node centrality degree.However,the computing speed and the calculation scale are limited.According to the characteristics of a protein network,the shortest path-based method of essential protein identification is presented.The method chooses reasonable identification thresholds and topology parameters to predict essential proteins of the whole protein-protein interaction network.Experiment results show that the proposed identification method can describe the importance of protein nodes effectively.The method can be used to identify the essential protein with good connectivity and big edge density of the whole protein-protein interaction network,without affecting the accuracy.