膀胱癌m6A相关长链非编码RNA预后模型的构建和免疫分析

Construction of a model for bladder cancer prognosis and immune analysis based on m6A-related long non-coding RNAs

目的:探究膀胱癌中N6-甲基腺苷(N6-methyladenosine,m6A)相关长非编码RNA(long non-coding RNA,lncRNA)及其与患者预后的关系。方法:通过癌症基因组图谱(the cancer genome Atlas,TCGA)数据库获取转录、突变和临床数据,获得m6A相关的lncRNA表达数据,采用共表达网络分析、Cox回归分析和最小绝对收缩和选择算法(least absolute shrinkage and se-lection operator,LASSO)回归分析鉴定与m6A相关的lncRNA,并构建风险预后模型和对其进行验证。然后,本研究还构建了列线图来预测膀胱癌患者的预后。通过基因本体(gene ontology,GO)、京都基因与基因组百科全书(kyoto encyclopedia of genes and genomes,KEGG)研究生物学功能的差异。使用ESTIMATE算法计算基质、免疫评分。采用单样本基因集富集分析(single sample gene set enrichment analysis,ssGSEA)算法定量分析免疫细胞浸润和免疫功能。采用oncoPredict算法预测潜在治疗药物。结果:本课题组确定了1739个m6A相关lncRNA,其中14个与预后相关。再通过LASSO回归分析筛选了5个m6A相关的lncRNA(GRK5-IT1、AC008883.2、AC145207.5、AC103746.1、AC104564.3)来构建预后预测模型。Kaplan-Meier和受试者工作特征曲线(receiver operating characteristic,ROC)曲线显示,该特征在训练集、验证集和全集中具有良好的预测能力。与其他临床特征相比,m6A相关lncRNA模型具有更高的诊断效率。免疫细胞浸润和ssGSEA分析进一步证实,特征lncRNA与膀胱癌患者的免疫状态显著相关。此外,高风险组对多种药物的治疗反应与低风险组存在显著差异。结论:5个m6A相关的特征ln-cRNA可能有助于评估膀胱癌患者的预后和免疫特征,并指导个性化治疗方案的制定。

Objective:To explore N6-methyladenosine(m6A)-related long non-coding RNAs(lncRNAs)in bladder cancer(BCa)and their correlations with prognosis.Methods:Transcription,mutation,and clinical data of m6A-related lncRNAs were downloaded from The Cancer Genome Atlas(TCGA)database.Then m6A-related signature lncRNAs were identified by co-expression network analysis,Cox regression analysis,and LASSO regression analysis.A risk prognostic model was constructed and validated.A nomogram was constructed to predict the prognosis of BCa patients.Functional enrichment analysis was performed using GO and KEGG.Stromal and immune scores were evaluated using the ESTIMATE algorithm.Single sample gene set enrichment analysis algorithm was used for quantitative analysis of immune cell infiltration and immune functions.The oncoPredict algorithm was used to predict potential thera-peutic drugs.Results:We identified 1,739 m6A-related lncRNAs,of which 14 were correlated with prognosis.Five m6A-related ln-cRNAs(GRK5-IT1,AC008883.2,AC145207.5,AC103746.1,and AC104564.3)were selected using LASSO regression to construct a prognostic risk model.The Kaplan–Meier and receiver operating characteristic curves confirmed that the prognostic risk model had good predictive performance in training,test,and complete sets.The prognostic risk model based on m6A-related lncRNAs had higher diagnostic efficiency compared to models based on other clinical fea-tures.Immune cell infiltration and single sample gene set enrich-ment analysis further confirmed that these signature lncRNAs were significantly associated with the immune status of BCa patients.Fur-thermore,the high-risk group and the low-risk group showed sig-nificant differences in the responses to multiple drugs.Conclusion:The five m6A-related signature lncRNAs may help assess the prog-nosis and immune characteristics of BCa patients and facilitate the determination of personalized treatment options.