基于图元向量的差异共表达分析研究

差异分析对于揭示生命体的生长、发育和衰老过程及疾病发生具有重大的意义,基于网络的差异分析方法已经成为系统生物学的一个研究热点.Przulj提出的图元及图元向量作为一描述网络局部结构信息的方法,已经在网络分析方法面取得了很多重要的结果.本文在图元向量的基础上提出了二种节点变化的差距度量方法,通过聚类可以分别挖掘网络中模块内变化基因簇和模块间变化基因簇.应用AGEMAP数据库中12个小鼠组织基因表达数据的结果表明:大部分聚类簇都高度显著富集与衰老相关的GO条目.

Mining Differential Co-expression Clusters Based on Graphlet Orbits

Differential analysis is a major method to understand the process of biological evolution and the progress of diseases.Recently, differential analysis based on graph has been a hot area in system biology.Since Przulj proposed the conception of graphlet and graphlet orbits, both of them have been used in various network analyses.In this paper, we proposed two measures based on graphlet orbits, and we used them to mine within module differential co-expresssion clusters and module to module differential co-expression clusters.Applications on the data of mice for 16 months and 24 months have shown that most of the clusters are significant enrichment in some GO terms related with aging.