肝细胞癌免疫相关基因和lncRNA联合预后模型的构建及验证

Construction and validation of a combined prognostic model of immune-related genes and lncRNAs in hepatocellular carcinoma

目的 构建肝细胞癌(HCC)免疫相关的基因(IRGs)和lncRNA(IRlncRNAs)联合预后模型。方法 2022年6-8月通过TCGA数据库下载HCC转录组及临床数据;对转录组中IRGs进行加权基因共表达网络分析(WGCNA)得到与预后相关的核心基因,对核心基因与转录组中lncRNA共表达分析得到IRlncRNAs;单因素Cox回归筛选生存相关的IRGs和IRlncRNAs,LASSO回归构建模型并进行验证;对高低风险病人差异表达的基因进行GO(基因本体论)和KEGG(京都基因与基因组百科全书)分析探索影响预后的可能机制。结果 构建了由6个IRGs及7个IRlncRNAs组成的预后风险模型;不同风险病人组织学分级(P=0.001)、临床分期(P=0.005)、T分期(P=0.010)差异有统计学意义;在训练集、测试集和总样本集中,高风险病人总生存期较低风险病人显著降低(均P<0.05);模型在预测HCC病人1年生存率中表现良好,训练集、测试集和总样本集受试者操作特征(ROC)曲线下面积分别为0.85、0.81和0.83;多因素Cox回归表明评分可独立于其他特征预测病人生存(P<0.001);GO分析显示差异基因主要参与有丝分裂等事件;KEGG分析显示差异基因参与了磷脂酰肌醇3激酶-蛋白激酶B、细胞周期等通路。结论基于IRGs和IRlncRNAs构建的HCC预后模型具有较好的预测价值,可能有助于HCC的临床决策和管理。

Objective To construct a combined prognostic model of immune-related genes(IRGs)and lncRNAs(IRlncRNAs)in he-patocellular carcinoma(HCC).Methods From June to August 2022,HCC transcriptome and clinical data from the Cancer Genome At-las(TCGA)database were downloaded.The weighted gene co-expression network analysis(WGCNA)of IRGs in the transcriptome was performed to obtain the core genes related to prognosis,and the co-expression analysis of core genes and IRlncRNAs in the transcrip-tome was performed to obtain IRlncRNAs.Univariate Cox regression analysis was used to screen survival-related IRGs and IRln-cRNAs,and LASSO regression was used to construct the combined risk model and the model was validated.Gene Ontology(GO)en-richment analysis and Kyoto Encyclopedia of Genes and Genomes(KEGG)signaling pathway analysis of differentially expressed genes between high and low risk patients were performed to explore the possible mechanism affecting prognosis.Results A prognostic model consisting of 6 IRGs and 7 IRlncRNAs was constructed.There were significant differences in histological grade classification(P=0.001),clinical stage(P=0.005),and T stage(P=0.010)among patients with different risks.The overall survival time of high-risk group was significantly reduced compared with that of low-risk group in train set,test set and total sample set(all P<0.05).The model per-formed well in predicting 1-year survival rate of HCC patients,and the areas under the ROC curve(AUC)in train set,test set,and total sample set were 0.85,0.81,and 0.83,respectively.Multivariate Cox regression showed that the score could predict patient survival in-dependently of other characteristics to predict patient survival(P<0.001).GO analysis showed that differential genes were mainly in-volved in mitosis and other events.KEGG analysis showed that differential genes were involved in PI3K-AKT,cell cycle and other path-ways.Conclusion The HCC prognostic model constructed based on IRGs and IRlncRNAs has a good predictive value and may help in HCC clinical decision-making and management.