目的: 探讨HER-2相关基因(the human epidermal growth factor receptor-2-related genes, HRGs)与膀胱癌生存预后的相关性，并基于HRGs构建一种膀胱癌患者生存预后的预测模型。方法: 从癌症基因组图谱(the cancer genome atlas, TCGA)中下载膀胱肿瘤组织mRNA测序数据和临床数据，通过与分子签名数据库(the molecular signatures database, MSigDB)中HER-2相关的基因联合分析鉴定膀胱癌中的HRGs。利用单因素和多因素Cox回归分析进一步明确与膀胱癌生存相关的HRGs(P < 0.05)，并构建HRGs风险模型(HRGs risk score model, HRSM)，根据风险评分取中位数将膀胱癌患者分成高风险组和低风险组。利用R语言对高风险和低风险组的患者进行生存分析，并对HRGs与临床特征的相关性进行分析。利用多因素Cox回归分析，验证影响膀胱癌患者预后的独立因素。计算HRSM的受试者工作特征曲线(receiver operating characteristic curve, ROC)下的面积(area under the curve, AUC)，并构建诺模图(nomogram)对膀胱癌患者进行生存预测。利用TIMER数据库对HRSM和患者免疫细胞浸润相关性进行分析。结果: 共鉴定到13个与患者生存相关的HRGs。通过多因素Cox回归分析，筛选出5个基因(BTC、CDC37、EGF、PTPRR和EREG)构建HRSM，高风险组的膀胱癌患者5年生存率明显低于低风险组患者。通过临床相关性分析发现，PTPRR的高表达与肿瘤分级、分期呈显著负相关，而EREG的高表达与肿瘤分级、分期呈正相关；EGF表达量的增加和患者的高级别有相关性，而CDC37的高表达却呈现出了相反的结果；BTC的表达与临床特征无显著相关性。通过对HRSM与免疫细胞的相关性分析发现，风险评分与树突状细胞、CD8⁺T细胞、CD4⁺T细胞、中性粒细胞和巨噬细胞的浸润呈正相关。结论: HRGs对膀胱癌患者的预后有重要作用，可能作为新的预测性生物标志物和治疗的潜在靶点。

Objective: To investigate the correlation between the human epidermal growth factor receptor-2-related genes (HRGs) and survival prognosis of bladder cancer and to construct a predictive model for survival prognosis of bladder cancer patients based on HRGs. Methods: HRGs in bladder cancer were found by downloading bladder tumor tissue mRNA sequencing data and clinical data from the cancer genome atlas (TCGA), downloading HER-2 related genes from the molecular signatures database (MsigDB), and crossing the two databases. Further identifying HRGs associated with bladder cancer survival (P < 0.05) by using single and multi-factor Cox regression analysis and constructing HRGs risk score model (HRSM), the bladder cancer patients were categorized into high-risk and low-risk groups accor-ding to the median risk score. Survival analysis of the patients in high- and low-risk groups was conducted using R language and correlation of HRGs with clinical characteristics. A multi-factor Cox regression analysis was used to verify the independent factors affecting the prognosis of the patients with bladder cancer. The area under the curve (AUC) of the receiver operating characteristic curve (ROC) of HRSM was calculated, and a nomogram was constructed for survival prediction of the bladder cancer patients. Analysis of HRSM and patient immune cell infiltration correlation was made using the TIMER database. Results: A total of 13 HRGs associated with patient survival were identified in this study. Five genes (BTC, CDC37, EGF, PTPRR and EREG) were selected for HRSM by multi-factor Cox regression analysis. The 5-year survival rate of the bladder cancer patients in the high-risk group was significantly lower than that of the patients in the low-risk group. High expression of PTPRR was found to be significantly and negatively correlated with tumor grade and stage by clinical correlation analysis, while EREG was found to be the opposite; Increased expression of EGF was associated with high grade, however, the high expression ofCDC37showed the opposite result. And no significant correlation was found between BTC expression and clinical features. Correlation analysis of HRSM with immune cells revealed a positive correlation between risk score and infiltration of dendritic cells, CD8⁺T cells, CD4⁺T cells, neutrophils and macrophages. Conclusion: HRGs have an important role in the prognosis of bladder cancer patients and may serve as new predictive biomarkers and potential targets for treatment.