基于COVID-19相关基因的肝细胞癌分子分型及预后模型构建与验证

Molecular typing of hepatocellular carcinoma based on COVID-19 related genes and construction and validation of prognostic model

目的:基于新型冠状病毒肺炎(COVID-19)相关基因对肝细胞癌进行分子分型,并建立预后模型。方法:收集GeneCard数据库中COVID-19相关基因,分析其在肝细胞癌患者癌与癌旁组织的基因表达量差异,通过一致聚类分析进行分子分型。使用limma包计算不同分子亚型之间差异表达基因并进行功能富集分析。使用R软件包“ESTIMATE”评估不同分子亚型的免疫评分。使用Cox比例风险回归模型和lasso回归构建多基因预后模型,并在其他数据库进行外部验证。结果:基于28个COVID-19相关基因将TCGA的365个肝细胞癌(LIHC)样本分成3个亚型,预后较差的C2亚型具有更高的免疫评分。多因素Cox回归分析结果说明5基因模型是肝细胞癌患者的独立预后危险因素。细胞周期、同源重组等肿瘤相关通路随着风险评分的升高而增加,提示高风险组具有更强的侵袭性。结论:构建了基于COVID-19相关基因的肝细胞癌分子亚型,并开发了预后相关的5基因模型(VEGFA、CD14、CD209、REN、PSMD1)。

Objective:To conduct molecular typing of hepatocellular carcinoma based on corona virus disease 2019(COVID-19)related genes and establish a prognostic model.Methods:COVID-19 related genes in the GeneCard database were collected,then the difference in gene expression between cancer and adjacent cancer in patients with hepatocellular carcinoma was analyzed,and molecular typing was performed through consensus clustering analysis.Differentially expressed genes between different molecular subtypes were calculated using the limma package and functional enrichment analysiswas performed.The immune scores of different molecular subtypes were evaluated using the R package"ESTIMATE".A multigene prognostic model was constructed using Cox proportional hazards regression model and lasso regression,and the model was externally validated in other databases.Results:Based on 28 COVID-19-related genes,the 365 liver cell carcinoma(LIHC)samples from TCGA were divided into three subtypes.The C2 subtype,which had a worse prognosis,had a higher immune score.The results of multivariate Cox regression analysis showed that the 5-gene model was an independent prognostic risk factor for patients with liver cancer.Cell cycle,homologous recombination,and other tumor-related pathways increased with the increase of risk score,indicating that the high-risk group was more aggressive.Conclusion:Molecular subtypes for hepatocellular carcinoma based on COVID-19 related genes are constructed and a prognostic 5-gene model(VEGFA,CD14,CD209,REN,PSMD1)is developed.