Abstract: Objective To construct a prognostic model for glioma patients based on cuproptosis related genes. Methods The transcriptome expression profiles of normal brain tissues were downloaded from GTEx database. The transcriptome expression profiles and corresponding clinical data of glioma patients were downloaded from TCGA and GEO databases. Differential expression analysis and prognostic analysis were performed on the 12 CRGs. Firstly, according to the expression profile of prognostic CRGs, the glioma samples were classified into different cuproptosis subtypes by consensus clustering algorithm. Prognostic DEGs between different cuproptosis subtypeswere obtained. Then, Lasso Cox regression analysis was used to construct the prognostic risk score model, and ROC curve was used to evaluate the predictive performance of this model. Finally, nomogram was constructed to improve the clinical applicability of the prognosis model. Results A total of 720 glioma samples were obtained. Using the consensus clustering algorithm, glioma samples were classified into two cuproptosis subtypes, CRGcluster A and CRGcluster B. A total of 1831 differentially expressed genes related to prognosis were identifiedbetween the two cuproptosis subtypes. A prognostic risk score model with eight genes was constructed in the train set and validated in the test set. ROC curve analysis showed that the AUC of the model in predicting the 1-, 3- and 5-year survival rates of glioma patients in the test set were 0.881, 0.894 and 0.874. According to the prognosis model, glioma samples were classified into high-risk group and low-risk group, and patients in the high-risk group had lower survival rate and shorter survival time （P<0.05）. Conclusion In this study, a prognostic risk score model for glioma was constructed based on cuproptosis related genes. This model can accurately predict the prognosis of patients with glioma, which provides a new way to evaluate the prognosis of patients with glioma.

Key words: glioma, cuproptosis, prognosis

摘要： 目的 基于铜死亡相关基因构建胶质瘤患者的预后模型。方法 从GTEx数据库下载正常脑组织转录组表达谱数据,从TCGA和GEO数据库下载胶质瘤患者的转录组表达谱和相应的临床数据,对12个铜死亡相关基因（CRGs）进行差异分析和预后分析。首先,根据预后相关CRGs的表达谱,利用一致性聚类算法对胶质瘤样本进行铜死亡分型,获得分型间预后相关的差异表达基因。接着,利用Lasso Cox回归分析构建预后风险评分模型并通过ROC曲线评估模型的预测性能。最后,构建诺莫图提高风险评分的临床适用性。结果 共获得720个胶质瘤样本。利用一致性聚类算法,胶质瘤样本分为两个铜死亡分型（CRGcluster A和CRGcluster B）,分型间共有1831个预后相关的差异基因。在训练集中构建了一个包含8个基因的预后风险评分模型,并在测试集中进行了验证。ROC曲线分析表明该模型在测试集中预测胶质瘤患者1、3、5年生存率的AUC分别为0.881,0.894,0.874。根据模型胶质瘤样本分为高风险组和低风险组,高风险组患者生存率较低,生存时间较短（P<0.05）。结论 本研究基于铜死亡相关基因构建了胶质瘤预后风险评分模型,具有预测胶质瘤患者预后的作用,为评估胶质瘤患者预后提供了新的途径。

关键词: 胶质瘤, 铜死亡, 预后