基于自噬和铁死亡基因的肾透明细胞癌预后模型的建立和验证

Establishment and verification of a renal clear-cell carcinoma tumor prognosis-model based on autophagy and ferroptosis genes

目的:建立基于自噬和铁死亡相关基因的预后模型,并基于肾透明细胞癌(ccrCC)的自噬基因(autophagy related genes,ARGs)和铁死亡基因(ferroptosis related genes,FRGs)模型评估预后。方法:通过癌症基因组图谱(TCGA)数据库中的ccRCC数据集识别与风险相关的ARGs和FRGs,进行功能富集和肿瘤分型分析,通过单变量和多变量Cox回归建立537例患者的预后风险模型。多指标ROC用于评估模型的准确性。最后使用GSE29609数据集验证。结果:共发现37个差异表达的基因。单变量和多变量Cox回归确定了8个与OS相关的风险相关基因:CASP4、PRKCQ、BNIP3、BAG1、BIRC5、CHAC1、ATG16L2、EIF4EBP1。Kaplan-Meier生存分析显示,高危组患者生存率较低,多指标ROC曲线下面积>0.75,说明模型预测准确率较高。然后基于CIBERSORT算法进行免疫细胞浸润评估。结论:基于8个ARGs和FRGs的肾透明细胞癌相关基因预后模型具有一定的准确性,可更准确地指导临床治疗。

Objective:To investigate the influence of autophagy-relevant genes(ARGs)and ferroptosis-related genes(FRGs)on the prognosis of patients with ccRCC(clear cell renal carcinoma),and evaluate the prognosis based on ARGs and FRGs models of ccRCC.Methods:Risk-related ARGs and FRGs were identified in the ccRCC dataset from the TCGA database,and functional enrichment analysis and tumor classification were performed.The Cox proportional hazards model of 37 ccRCC patients was established through univariate and multivariate Cox regression.Multi-index ROC curve was used to assess the accuracy of the model.Finally,a GSE29609 dataset was used to verify whether our conclusion is feasible or not.Results:A total of 37 differentially expressed ARGs and FRGs were found.Univariate and multivariate Cox regression identified 8 risk-related genes related to OS,which are CASP4,PRKCQ,BNIP3,BAG1,BIRC5,CHAC1,ATG16L2 and EIF4EBP1.A Kaplan-Meier analysis results showed that patients at high risk of morbidity had lower survivability.The area(performance index)under the ROC curve was>0.75,indicating that the model’s prediction accuracy was higher.Then the immune cell infiltration were evaluated based on the CIBERSORT algorithm.Conclusion:The prognostic model of ccRCC-related genes based on 8 ARGs and FRGs has a certain accuracy and could be used to guide clinical treatment more accurately.