Combined gene expression and protein interaction analysis of dynamic modularity in glioma prognosis
Combined gene expression and protein interaction analysis of dynamic modularity in glioma prognosis
复制标题
DOI：
10.1007/s11060-011-0757-4
复制
发表时间：
2012-04-01
影响因子：
3.9
通讯作者：
Yang, Lizhuang
中科院分区：
文献类型：
--
作者：
Zhang, Xiaoyu;Yang, Hongbin;Yang, Lizhuang
由于影响胶质瘤预后的因素多种多样，预测患者的生存尤其困难。蛋白质-蛋白质相互作用(PPI)网络被认为是关于它们的空间特征如何与胶质瘤相关的。然而，体内PPI的动态性质使得它们在时间和空间上都是复杂的事件。预后特异性共表达信息的整合进一步增加了这些网络的动态特征。虽然已经确定了一些预测胶质瘤预后的生物标志物，但没有一种生物标志物足以准确预测预后或改善生存率。我们已经建立了共表达的蛋白质相互作用网络，将蛋白质-蛋白质相互作用与与不同生存时间相关的胶质瘤基因表达谱整合在一起。通过对胶质瘤预后网络，特别是亚网络的动态特征的对比分析，确定与胶质瘤预后相关的生物标志物。随着寿命的延长，4个显著差异表达基因(SDEGs)上调，10个SDEGs下调。此外，97个增强的差异共表达蛋白相互作用(DCPI)和99个减弱的DCPI与胶质瘤患者的寿命延长有关。我们提出了一种基于动态模块化网络构建的神经胶质瘤预后评估方法。我们已经使用这种方法来识别与胶质瘤预后相关的动态基因和相互作用。其中，MYC表达增强与寿命延长有关，E2F1与RB1、EGFR与p38之间的相互作用也与此有关。该方法为研究决定胶质瘤预后的分子机制提供了一种新的手段。
Because of the variety of factors affecting glioma prognosis, prediction of patient survival is particularly difficult. Protein-protein interaction (PPI) networks have been considered with regard to how their spatial characteristics relate to glioma. However, the dynamic nature of PPIs in vivo makes them temporally and spatially complex events. Integration of prognosis-specific co-expression information adds further dynamic features to these networks. Although some biomarkers for glioma prognosis have been identified, none is sufficient for accurate prediction of either prognosis or improved survival. We have established co-expressed protein-interaction networks that integrate protein-protein interactions with glioma gene-expression profiles related to different survival times. Biomarkers related to glioma prognosis were identified by comparative analysis of the dynamic features of the glioma prognosis network, particularly subnetworks. Four significantly differently expressed genes (SDEGs) are upregulated and ten SDEGs downregulated as lifetime is extended. In addition, 97 enhanced differently co-expressed protein interactions (DCPIs) and 99 weakened DCPIs were associated with glioma patient lifetime extension. We propose a method for estimating glioma prognosis on the basis of the construction of a dynamic modular network. We have used this method to identify dynamic genes and interactions related to glioma prognosis. Among these, enhanced MYC expression was related to lifetime extension, as were interactions between E2F1 and RB1 and between EGFR and p38. This method is a novel means of studying the molecular mechanisms determining prognosis in glioma.