基于贝叶斯准则的差异网络分析方法研究与应用  

Differential Gene Regulatory Network Inference Method Based on Bayesian Model Selection and its Application

作　　者：宋微 王文杰[1] 徐欢 蔡雨晴 李康[1] Song Wei;Wang Wenjie;Xu Huan(Department of Medical Statistics,Harbin Medical University (150081),Harbin)

机构地区：[1]哈尔滨医科大学卫生统计学教研室

出　　处：《中国卫生统计》2019年第4期484-488,共5页Chinese Journal of Health Statistics

基　　金：国家自然科学基金(81773551)

摘　　要：目的探讨基于贝叶斯准则的差异网络分析方法的性能,并将其应用于卵巢癌基因表达谱数据分析。方法通过模拟实验评价其识别差异边及差异节点的准确性,并与传统方法做对比。同时应用上皮性卵巢癌基因组学数据,构建差异网络模型。结果模拟试验结果表明,基于贝叶斯准则的差异网络分析方法识别差异边能力明显优于高斯图模型方法;实例分析结果表明,本文方法构建的差异网络模型具有实际意义。结论应用基于贝叶斯准则的差异网络分析方法能得出准确度较高的差异网络,效果优于传统方法。Objective To explore the performance of differential gene regulatory network inference method baseds on Bayesian model selection,and to apply it to the analysis of ovarian cancer gene expression data.Methods The efficiency of the algorithm given in this article was testified with simulation data and compared with traditional methods,and this algorithm was used to construct differential gene regulatory networks with ovarian cancer gene expression data.Results Differential gene regulatory network inference method based on Bayesian model selection performed better than the general Gaussian graphical models in differential edges inference.The results of gene expression data analysis also indicated that this algorithm could provide valuable differential network structures.Conclusion Differential gene regulatory network inference method based on Bayesian model selection can establish highly precise differential network models, performed better than traditional methods.

关 键 词：贝叶斯模型选择 高斯图模型 差异网络 

分 类 号：R737.31[医药卫生—肿瘤]