肺腺癌中综合分析竞争性内源RNA网络预测潜在的预后lncRNA

Comprehensive analysis of competitive endogenous RNA network reveals potential prognostic lncRNA in lung adenocarcinoma

目的:利用生物信息学构建竞争性内源RNA(ceRNA)网络进而探索肺腺癌的发病机制,为肺腺癌的预防和治疗提供生物靶标。方法:从癌症基因组图谱(TCGA)数据库下载肺腺癌组织和癌旁组织的基因表达谱。利用R软件获得与肺腺癌相关的差异表达RNAs(lncRNA、miRNA、mRNA),筛选标准为|LogFC|>1,P<0.01。miRcode数据库预测差异lncRNA的目标基因,运用miRDB、miRTarBase、TargetScan数据库预测miRNA的靶基因,构建ceRNA网络。进一步对ceRNA网络中的差异mRNA进行功能和通路富集分析,并对ceRNA网络中的lncRNA进行单因素和多因素分析,从而建立风险模型。R软件对特定的临床信息进行单因素和多因素分析,获得可以作为独立预后因子的临床信息。R软件对lncRNA和m RNA的相关性进行分析。Kaplan-Meier数据库对PPI网络中与lncRNA显著相关的关键基因进行生存分析。结果:从TCGA数据库中下载了535个肺腺癌组织样本和59个癌旁组织样本。R软件分析获得4847个差异mRNA、2070个差异lncRNA和232个差异miRNA。发现差异基因的主要功能为调控RNA聚合酶Ⅱ特异性、DNA结合转录激活剂活性、L-谷氨酸跨膜转运蛋白活性、Wnt信号等。获得由170个lncRNA、14个miRNA和66个mRNA组成的ceRNA网络。单因素分析得到7个与肺腺癌生存期显著相关的lncRNA,多因素分析最终得到6个可以作为独立预后因子的lncRNA(DGCR5、LINC00536、ATG9B、FGF14-IT1、ANO1-AS2、SYNPR-AS1)和风险值(1.245336)。在单因素分析中临床信息:肿瘤分期(Stage)、淋巴结(N)、远处转移(T)作为预后因子,多因素分析临床信息不可作为独立预后因子。在肺腺癌中,m RNA(FGF2、CCNB1、ALDOA、COL1A1、MCM4)低表达组的生存率高于高表达组。结论:本研究中的ceRNA网络着眼于疾病阶段和差异基因表达,可为ceRNA网络在肺腺癌发生和发展中的作用提供有价值的见解。

Objective:To analyze the competitive endogenous RNA(ceRNA)network of lung adenocarcinoma by bioinformatics and explore its pathogenesis,in order to provide biological targets for the prevention and treatment of lung adenocarcinoma.Methods:Gene expression profiles of lung adenocarcinoma tissues and paracancerous tissues were downloaded from The Cancer Genome Atlas(TCGA)database.R language-related program package was used to analyze differentially expressed RNAs(lncRNAs,miRNAs,mRNAs)(|LogFC|>1,P<0.01).The miRcode database was used to predict the target genes of differential lncRNAs,and miRDB,miRTarBase and TargetScan databases were used to predict target genes of differential miRNAs,and the ceRNA network was constructed.Functional and pathway enrichment analysis was performed on m RNAs in the ceRNA network.Univariable and multivariable analyses were performed on lncRNAs in the ceRNA network to establish a risk model.Univariable and multivariable analyses of specific clinical information was also performed to predict independent prognostic factors.The correlation between lncRNA and mRNA was analyzed by R software.Survival analysis was performed on key genes significantly associated with lncRNAs in the PPI network using the Kaplan-Meier database.Results:A total of 535 lung adenocarcinoma tissue samples and 59 paracancerous tissue samples were downloaded from the TCGA database.And 4847 differential mRNAs,2070 differential lncRNAs and 232 differential miRNAs were obtained by R software analysis.The main functions of differential genes were regulation of RNA polymeraseⅡspecificity,DNA binding transcription activator activity,L-glutamate transmembrane transporter activity,Wnt signal and so on.A ceRNA network consisting of 170 lncRNAs,14 miRNAs and 66mRNAs was constructed.Univariable analysis identified 7 lncRNAs that were significantly related to survival time of lung adenocarcinoma,and further multivariable analysis identified 6 lncRNAs(DGCR5,LINC00536,ATG9B,FGF14-IT1,ANO1-AS2,SYNPR-AS1)that could be used as independent prognosis factors,and the risk value was 1.245336.Univariable analysis of clinical information revealed that tumor stage,lymph node,and distant metastasis were prognostic factors,while multivariable analysis showed that they could not be used as independent prognostic factors.The survival rate of the low expression group of FGF2,CCNB1,ALDOA,COL1A1 and MCM4 mRNA was higher than that of the high expression group in lung adenocarcinoma.Conclusion:The ceRNA network in this study focuses on the disease stage and differential gene expression,which may provide valuable insights into the role of ceRNA network in the occurrence and development of lung adenocarcinoma.