论文部分内容阅读
目的:构建肺腺癌预后诊断模型并挖掘模型诊断价值。方法:数据来源于癌症基因组图谱(TCGA)，搜索截止日期为2020年4月10日。通过似然比检验比较肺腺癌组织(n n＝515)和癌旁组织(n n＝46)，筛选差异表达微小RNA(miRNA，miR)，单因素比例风险回归模型(COX回归)分析差异miRNA与预后相关性，鲁棒性分析找到可靠性最高miRNA。以多因素COX回归分析构建风险评分模型。在受试者工作特征(ROC)曲线找到最佳截止点后将患者分为高低风险组，生存曲线判断模型诊断效果。n 结果:得到差异miRNA 257个，其中12个miRNA与患者预后相关性较强(n P<0.05)。12-miRNA风险评分模型可发现组内高风险(n n＝131)患者生存时间较低风险(n n＝130)患者短[风险比(n HR)＝0.25、0.08、0.25、0.04、0.09、0.004、0.26、0.22、0.08、0.01、0.02、-0.09，n χ2＝13.85、11.30、7.54、6.05、5.76、5.79、4.85、4.74、4.66、4.51、4.25、4.22，n P<0.05]。用最佳截止点1.322重新分组(高风险组为80、95、176例；低风险组为180、166、346例)，高风险组生存期仍低于低风险组(高风险＝15.5、31.7、29.9个月，低风险＝59.7、55.1、59.7个月，n χ2＝32.80、8.40、28.20，n P<0.01，n P值均<0.01)。n 结论:12-miRNA风险评分模型可以为肺腺癌诊断提供新途径。“,”Objective:To construct a diagnostic model for prognosis of lung adenocarcinoma and digging the diagnostic value of the model.Methods:Data was downloaded from the Cancer Genome Atlas. The search deadline is April 10, 2020. Differentially expressed micro-RNAs (DEMs) between lung adenocarcinoma tissues (n n=515) and adjacent tissues (n n=46) were screened out by using likelihood ratio test. The correlation between DEMsand prognosis of patients by univariate Cox regression analysis and robust likelihood-based survival analysis was used to find the most reliable microRNA (miRNAs, miR). Multivariate Cox regression analysis was used to construct a risk-score model. After finding the best cut-off point on the receiver operating characteristic (ROC) curve, the patients were divided into high-and low-risk groups, and the survival curves were carried out to judge the diagnostic effect of the model in different data sets.n Results:257 DEMs were screened out. A total of 12 miRNAs which were strongly correlated with the prognosis of patients were obtained [hazard ratio (n HR)=0.25, 0.08, 0.25, 0.04, 0.09, 0.004, 0.26, 0.22, 0.08, 0.01, 0.02, -0.09, n χ2=13.85, 11.30, 7.54, 6.05, 5.76, 5.79, 4.85, 4.74, 4.66, 4.51, 4.25, 4.22, n P<0.05]. The 12-miRNA risk-score model indicated the survivaltime of patients in high risk group (n n=131) is shorter than patients in low risk group (n n=130), which was 37.8 months and 77.8 months, respectively (n χ2=19.70, n P<0.01). Using the the optimal cut-off point of 1.322 as the standard, the training group, the testing group and the overall group were divided into high (n n=80, 95, 176) and low (n n=180, 166, 346) risk groups respectively. The 12-miRNA model can still distinguish the survival time of each group (high risk=15.5, 31.7, 29.9 months, low risk= 59.7, 55.1, 59.7 months, n χ2=32.80, 8.40, 28.20, alln P<0.01).n Conclusion:The 12-miRNA risk-score model can provide a new way for the diagnosis of lung adenocarcinoma.