基于因果推断方法检验mRNA在年龄和低级别胶质瘤中的中介效应  

Assessing Age Effect on Low-grade Glioma Mediated by mRNA based on a Causal Inference Test

作　　者：王碧珏 郭志飞 杨海涛[3] 李治 王菊平[5] 曹红艳[5,6] 周立业 Wang Bijue;Guo Zhifei;Yang Haitao(Department of Social Medicine,Shanxi Medical University(030001),Taiyuan)

机构地区：[1]山西医科大学社会医学教研室,030001 [2]山西医科大学卫生管理学教研室 [3]河北医科大学流行病与卫生统计学教研室 [4]中北大学体育学院 [5]山西医科大学卫生统计教研室 [6]重大疾病风险评估山西省重点实验室

出　　处：《中国卫生统计》2020年第6期840-843,847,共5页Chinese Journal of Health Statistics

基　　金：国家自然科学基金资助(71403156,81872717);山西省回国留学人员科研资助(2017-054);山西省应用基础研究计划(201901D111204)。

摘　　要：目的构建年龄、mRNA表达水平和低级别胶质瘤(LGG)患者结局之间的因果中介推断模型,识别介导年龄和LGG结局的中介因子,为LGG的治疗和预后提供方法指导。方法以TCGA数据库中LGG数据为例,采用CIT和VanderWeele方法建立因果中介模型,识别年龄和LGG结局之间的介导基因。结果识别出11个在年龄和LGG结局之间起中介作用的基因,中介间接效应的RR值均大于1,表明这11个基因的表达异常可能是LGG不良结局的危险因素。结论基于因果推断的方法能够识别介导年龄和低级别胶质瘤结局的重要基因,为LGG的预后治疗提供了潜在的预后标志物和治疗靶点。Objective A causal mediating inference model between age,mRNA expression level and outcome of low-grade glioma patients was constructed to identify mediating factors that mediate age and the outcome of LGG,providing methodological guidance for the treatment and prognosis of LGG.Methods Taking LGG data in TCGA database as an example,CIT method and VanderWeele method were used to establish causal mediation model to identify the mediated genes between age and LGG outcome.Results Eleven genes mediating between age and LGG outcome were identified,and the RR values of the indirect effects of these mediations were all greater than 1,indicating that abnormal expression of these 11 genes may be risk factors for adverse LGG outcomes.Conclusion The method based on causal inference could identify important genes that mediate age and low-grade glioma outcomes,providing potential prognostic markers and therapeutic targets for LGG prognosis.

关 键 词：因果推断 中介分析 MRNA 低级别胶质瘤 

分 类 号：R195.1[医药卫生—卫生统计学]