基于网络资本评估和节点失效的关键蛋白质识别算法研究  

Identification of essential proteins based on network capital assessment and invalidating protein node

作　　者：许睿[1] 胡萍[1] 李纲[1] 

机构地区：[1]河南科技学院,河南新乡453003

出　　处：《智能计算机与应用》2016年第6期25-28,共4页Intelligent Computer and Applications

摘　　要：针对目前蛋白质节点删除方法存在破坏蛋白质相互作用网络拓扑结构的问题,在定义蛋白质节点可达性的基础上,提出了一种基于蛋白质网络资本评价和节点失效法的关键蛋白质节点识别算法。在保持网络整体拓扑结构完整的前提下,通过分析蛋白质节点失效前后网络资本值的下降程度评估该节点在蛋白质网络中的重要程度,从而识别关键蛋白质。实验结果表明本算法对最为重要的前1%和前5%数据的关键蛋白质预测结果优于四种典型的中心性测度方法,而且对于关键蛋白质的识别具有较高的准确性。The traditional elimination method of protein node has the problem of destroying network topology. This paper defines accessibility of protein nodes,and develops a new algorithm based on the network capital assessment and invalidating protein node.Under the premise of network topology integrity,this algorithm evaluates the importance of nodes and identifies essential proteins in Protein-Protein Interaction network by analyzing the declining extent of network capital before and after invalidating protein node. The experimental results show that this algorithm is better than four typical Centrality Measures on the identification of essential proteins in the Top 1% and Top 5% data,and has higher accuracy on the identification of essential proteins.

关 键 词：蛋白质网络 可达性 网络资本评估 关键蛋白质 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]