目的 采用生物信息技术对肺腺癌(LUAD)中m1A相关的长链非编码RNA(mRLs)进行分析并构建预后风险模型.方法 从癌症基因组图谱(TCGA)数据库中获得转录组数据和临床资料,通过对m1A相关基因和lncRNAs进行共表达分析确定mRLs.最终筛选得到472例临床样本,其中男218例,女254例;年龄35～78(59.4±23.2)岁.通过R语言中"caret"包的create Data Partition函数将样本分成训练集(236例)和测试集(236例).在训练集中采用单因素、多因素Cox分析、最小绝对收缩和选择算子回归分析构建预后风险模型,计算评分后按照其中位值分为高、低风险组,并使用受试者工作特征曲线(ROC)、Cox回归分析、主成分分析进一步评估该模型的准确性.对该模型高低风险组间的肿瘤突变负荷(TMB)和免疫细胞浸润进行分析.两组间比较通过Wilcoxon检验完成.结果 共筛选到146个mRLs,46个差异表达的mRLs,最终确定23个mRLs构成LUAD预后风险模型,该模型在训练集患者的1、3、5年生存预测性能曲线下面积(AUC)值分别为0.808、0.746和0.743;高风险组患者的总生存期明显低于低风险组(P＜0.01),测试集和全集中也呈现出良好的预测效能,进一步分析显示高风险组患者肿瘤微环境免疫浸润更多.结论 基于23个mRLs构建的预后风险模型能较好的预测LUAD患者的预后情况.
Establishment and evaluation of m1A-related long non-coding RNA prognostic risk model for lung adenocarcinoma
Objective Bioinformatics was used to analyze mlA-related long non-coding RNAs(mRLs)in lung adenocarcinoma(LUAD)and construct a prognostic risk model.Methods The transcrip-tome data and clinical data were obtained from the cancer genome atlas(TCGA)atabase,and mRLs were identified by co-expression analysis of m1A-related genes and lncRNAs.Finally,472 clinical samples were screened,including 218 males and 254 females.The age was 35-78(59.4±23.2)years old.Through the createDataPartition function of the"caret"package in R language,the samples were divided into training set(236 cases)and test set(236 cases).Univariate and multivariate Cox regression analysis and Least Abso-lute Shrinkage and Selection Operator(LASSO)regression analyses were used to construct a prognostic risk model in the training set.After calculating the score,they were divided into high and low risk groups ac-cording to their median values,and the receiver operating characteristic curve(ROC),Cox regression anal-ysis and principal component analysis were used to further evaluate the accuracy of the model.Tumor muta-tion burden(TMB)and immune cell infiltration between the high and low risk groups of the model were an-alyzed.The comparison between the two groups was completed by Wilcoxon test.Results A total of 146 mRLs,of which 46 mRLs were differentially expressed,and 23 mRLs were finally determined to con-stitute the LUAD prognostic risk model.The area under the curve(AUC)values of the model in the 1,3 and 5 year survival prediction performance curves of the patients in the training set were 0.808,0.746 and 0.743,respectively.The overall survival of patients in the high-risk group was significantly lower than that in the low-risk group(P＜0.01).The test set and the entire set also showed good predictive efficacy.Fur-ther analysis showed that patients in the high-risk group had more immune infiltration in the tumor microen-vironment.Conclusion The prognostic risk model based on 23 mRLs can better predict the prognosis of LU AD patients.
NETL
NSTL
万方数据
