摘要
与生物实验方法相比,基于网络拓扑的生物信息学方法在关键节点识别上有独特优势.基于网络拓扑的关键节点识别主要依赖节点在生物网络中的拓扑特性,通过观察节点网络拓扑参数的大小、所处的路径或模块的结构及其动力特性,在一定程度上可以对其关键性进行推断.从节点的中心性测度、网络的拓扑参数及层次结构等几方面总结了生物网络及其节点的主要拓扑特征;比较了蛋白质网络、代谢网络及基因调控网络关键节点识别的主要方法;分析了节点拓扑参数计算、路径求解及模块的划分及识别算法;指出生物网络关键节点识别上存在识别率不高、不同研究结论的不一致甚至相互矛盾、现有算法对网络规模日益增长的不适应等问题,并提出解决问题的思路及进一步研究的方向.
The essentiality of a node is correlated with its topological properties in a bionetwork. Comparing with other methods such as biological experiments, bioinformatics methods based on topology possess particular advantage in the identification of essential nodes. By investigation the topological parameters and the position in a bionetwork, the essentiality of a node can be predicted. This paper summarizes important characteristics both from bionetworks and their nodes, analyzes main techniques and algorithms relation to the identification from several kinds of bionetwork. While problems standing in the process are pointed out, such as low identification ratio, inconsistency between different researchers, and incommensurate of existing algorithms to huge amounts of computation deriving from the growing of bionetworks and the dynamic of identification course, several ideas are proposed to the solution.
出处
《数学的实践与认识》
CSCD
北大核心
2011年第7期114-125,共12页
Mathematics in Practice and Theory
基金
国家自然科学基金(60433020)
广西教育厅科研项目(2000911MS196)
关键词
网络拓扑
生物信息
关键节点
识别
复杂网络
network topology
bio information
essential nodes
identification
complex networks
