基于生物信息学方法构建肝癌患者预后相关内源竞争RNA调控网络

Bioinformatics-based Approach for the Construction of Prognosis-related Competing Endogenous RNA Regulatory Network in Patients with Hepatocellular Carcinoma

目的:基于生物信息学方法探索肝细胞癌相关的差异表达基因(DEGs),并构建预后相关内源竞争RNA(competing endogenous RNA,ceRNA)调控网络。方法:利用数据集GSE89377,筛选肝细胞癌不同发展时期的肿瘤组织与正常组织间的差异表达基因;通过GEPIA,HPA数据库探究DEGs的mRNA和蛋白在肝细胞癌中的表达,及对患者预后的影响;随后通过Cox回归分析,筛选可独立预测患者预后的关键基因;进一步通过GO和KEGG富集分析探索关键基因相关信号通路。最后,通过miRmap、miRWalk和Targetscan数据库预测关键基因的上游微小RNA (microRNA,miRNA),并筛选预后相关miRNAs;通过Starbase和Lncbase预测重要miRNAs的LncRNAs,并筛选预后相关LncRNAs。以此构建影响肝癌患者预后的差异基因-miRNA-LncRNAs的ceRAN调控网络。结果:韦恩分析及表达分析发现,4个DEGs在肝癌组织及正常组织间差异表达,即AKR1B10、LAGLS4、MUC13、IGFALS;其中,AKR1B10高表达可独立预测肝癌患者不良结局;共预测到5个AKR1B10的上游miRNA,其中miRNA-486-5p与AKR1B10的表达呈显著负相关,其低表达可降低肝癌患者的总生存率;共预测到4个miRNA-486-5p的lncRAN,生存分析显示RAB30-AS1的表达影响肝细胞癌患者的预后。通过上述分析,构建了影响肝癌患者预后的ceRAN调控网络,即AKR1B10-miRNA-486-5p-RAB30-AS1。结论:通过生物信息学分析,构建了肝癌患者预后相关ceRAN调控网络AKR1B10-miRNA-486-5p-RAB30-AS1,有望成为肝癌治疗的新策略。

Objective:To explore hepatocellular carcinoma-associated differentially expressed genes(DEGs) based on a bioinformatics approach and construct a prognosis-associated competing endogenous RNA(ceRNA) network.Methods:The dataset GSE89377 was used to screen for differentially expressed genes between tumor tissue and normal tissue at different stages of hepatocellular carcinoma development.The expression of mRNA and protein of DEGs in hepatocellular carcinoma and the impact on patient prognosis were explored by GEPIA,HPA database,followed by cox regression analysis to screen key genes that can independently predict patient prognosis,and further by GO and KEGG enrichment analysis to explore key gene-related signaling pathways.The miRmap,miRWalk and Targetscan databases were used to predict upstream microRNAs(miRNAs) of key genes and screen for prognosis-related miRNAs,and Starbase and Lncbase were used to predict LncRNAs of important miRNAs and screen for prognosis-related LncRNAs.In this way the ceRAN regulatory network of differential gene-miRNA-LncRNAs affecting the prognosis of hepatocellular carcinoma patients was constructed.Results:Wayne analysis and expression analysis revealed that four DEGs were differentially expressed between liver cancer tissues and normal tissues,namely AKR1B10,LAGLS4,MUC13,and IGFALS.Among them,high AKR1B10 expression independently predicted poor outcome in patients with hepatocellular carcinoma.A total of five upstream miRNAs of AKR1B10 were predicted,among which miRNA-486-5p was significantly and negatively correlated with AKR1B10 expression,and its low expression reduced the overall survival rate of liver cancer patients.A total of four miRNA-486-5p lncRANs were predicted,and survival analysis showed that the expression of RAB30-AS1 affected the prognosis of patients with hepatocellular carcinoma.Through the above analysis,a ceRAN regulatory network,AKR1B10-miRNA-486-5p-RAB30-AS1,was constructed to affect the prognosis of hepatocellular carcinoma patients.Conclusion:The prognosis-related ceRAN regulatory network AKR1B10-miRNA-486-5p-RAB30-AS1 was constructed by bioinformatics analysis for liver cancer patients,which is expected to be a new strategy for liver cancer treatment.