乳腺癌预后的相关基因生物信息学分析

Bioinformation analysis of data regarding breast cancer prognosis in European and American populations

目的:通过对肿瘤基因组图谱计划(TCGA)基因表达谱和micro RNA测序数据的生物信息学分析,寻找乳腺癌预后相关的关键分子事件。方法:收集并整理TCGA数据库中乳腺癌基因表达谱和micro RNA测序数据,通过线性模型和经验贝叶斯方法结合传统t检验筛选乳腺癌中异常表达的基因和micro RNA,利用R语言包micro RNA-m RNA预测micro RNA潜在靶向调控的基因。另外通过基因集富集算法(GAGE)分析预测micro RNA及其潜在靶基因在乳腺癌中参与的异常信号调控通路;Cox回归风险模型探讨影响乳腺癌患者预后的关键分子事件。结果:共发现17 533个基因中有344个基因在乳腺癌中异常表达,同时鉴定了135个异常表达的micro RNA;通过预测分析发现8个micro RNA及31个潜在调控的靶基因参与了139条乳腺癌中异常变化的分子信号通路;Cox回归风险模型结果显示SFRP1表达升高提示预后良好(HR=0.9,P=0.015),has-mi R-342-5p单独作用对乳腺癌预后无明显影响(HR=0.99,P=0.144),然而,交互作用研究显示,has-mi R-342-5p抑制SFRP1表达时提示乳腺癌病人预后不良(HR=1.88,P=0.016)。结论:通过对TCGA乳腺癌基因表达谱和micro RNA测序数据深入挖掘发现了乳腺癌中表达异常的micro RNA及其靶基因,进一步分析了它们所参与的关键信号通路,并揭示has-mi R-342-5p抑制SFRP1表达的分子事件发生时,乳腺癌病人预后较差。

OBJECTIVE:To understand key molecular events in prognosis of breast cancer,we analyzed the data published by TCGA,and to study the associations between microRNAs and potential target genes at the pathway level.METHODS:To collect and organize the gene expression profile and microRNA sequencing data of breast cancer in TCGA database,we use multiple t-tests to analyze the differentially expressed microRNAs and target genes,and R language package microRNA-mRNA to predict the potential target regulated genes in microRNA.General Applicable Gene-set Enrichment(GAGE)was applied to discover the key genes in essential pathways of breast cancer.Integrated association analysis was used to find the potential targets of the microRNAs.The Cox proportional hazard model was used to evaluate the possible prognostic signatures in breast cancer.RESULTS:The results show that there were abnormal expression of 344 genes and 135 microRNAs.8 microRNAs with 31 potential target genes might have played key roles in breast cancer through 139 essential pathways.Cox regression risk model results show that increased SFRP1 had a protective effect(HR=0.9,P=0.015)while miR-342-5p had neither protective nor hazard effects(HR=0.99,P=0.144).However interactive analysis suggest that has-mir-342-5p inhibited SFRP1 expression in breast cancer with poor prognosis(HR=1.88,P=0.016).CONCLUSION:Through deep excavation of gene expression and microRNA sequencing data from TCGA database,we identified abnormal expression of certain microRNA and genes in breast cancer,and their involvement in key signaling pathways,suggesting that has-mir-342-5p inhibited the expression of SFRP1 in patients with breast cancer who had poor prognosis.