基于网络成分分析的阿尔茨海默症靶基因动态表达研究

Using Network Component Analysis to Dissect Dynamic Expression of Gene and Network Structure in Alzheimer's Disease

解析转录因子与靶基因之间相互调节的关系并构建转录调控网络,对研究阿尔茨海默症(AD)的致病机理、早期诊断及制药等具有重要意义。网络成分分析(NCA)是一种能够动态预测转录因子活性并表现其影响关系的方法。本研究利用转录因子对靶基因的调控作用,及基因在AD不同病程中表达的先验知识和生物数据,通过预处理AD基因表达数据,选择出10个转录因子和85个靶基因进行网络成分分析,并利用162条调控关系构建AD基因调控网络,形成和展示了转录因子对靶基因的动态调控关系和作用。通过动态预测转录因子活性及构建网络图,发现转录因子在AD疾病的不同程度的活性有明显变化趋势,其调控的靶基因变化符合AD的病理特征。如靶基因NONO在转录因子ANAPC5的调控下,表达值由3 126上升至4 508,而靶基因YWHAZ表达值由6 000下降到接近于0。该研究为AD致病机理探寻、早期诊断和相应的分子生物学实验,提供了新的思路和依据。

Regulating the relationship between the analysis of transcription factors and target genes and building transcriptional regulatory networks are of great significance for the study of Alzheimer ＇s disease （AD） pathogenesis, early diagnosis and pharmaceutical. Network component analysis （NCA） is a way to dynamically predict transcriPtion factor activity and performance affect relations. According to a priori knowledge that transcription factor regulates multiple genes and induces the changes of gene expression, in this article 10 transcription factors were pretreated and 85 target genes were selected for network component analysis, and 162 of regulation relationship were used to build AD gene regulatory networks to form and show the dynamic regulation of relations and the role of transcription factor target genes. Dynamic prediction of transcription factor activity was varied significantly in line with the changes of its target genes regulating the pathological features of AD. For example, the expression values of the target gene NONO was increased from 3 126 to 4 508 in control with TF ANAPC5, but the expression values of the target gene YWHAZ was decreased from 6 000 to close to 0. The study provided a new insight and theoretical basis for The pathogenic mechanism and early diagnosis of AD.