This study was conducted to identify genes that are differentially expressed in paracancerous tissue and to determine the potential predictive value of selected gene panel. Gene transcriptome data of bladder tissue was downloaded from UCSC Xena browser and NCBI GEO repository, including GTEx (the Genotype-Tissue Expression project) data, TCGA (The Cancer Genome Atlas) data, and GEO (Gene Expression Omnibus) data. Differentially Expressed Genes (DEGs) analysis was performed to identify tumor-DEGs candidate genes, using the intersection of tumor-paracancerous DEGs genes and paracancerous-normal DEGs genes. The survival-related genes were screened by Kaplan–Meier (KM) survival analysis and univariable Cox regression with the cutoff criteria of KM < 0.05 and cox p-value < 0.05. The risk model was developed using Lasso regression. The clinical data were analyzed by univariate and multivariate Cox regression analysis. Gene Ontology (GO) and KEGG enrichment analysis were performed in the DEGs genes between the high-risk and low-risk subgroups. We identified six survival-related genes, EMP1, TPM1, NRP2, FGFR1, CAVIN1, and LATS2, found in the DEG analyses of both, tumor-paracancerous and paracancerous-normal differentially expressed data sets. Then, the patients were classified into two clusters, which can be distinguished by specific clinical characteristics. A three-gene risk prediction model (EMP1, FGFR1, and CAVIN1) was constructed in patients within cluster 1. The model was applied to categorize cluster 1 patients into high-risk and low-risk subgroups. The prognostic risk score was considered as an independent prognostic factor. The six identified survival-related genes can be used in molecular characterization of a specific subtype of bladder cancer. This subtype had distinct clinical features of T (topography), N (lymph node), stage, grade, and survival status, compared to the other subtype of bladder cancer. Among the six identified survival-related genes, three-genes, EMP1, FGFR1, and CAVIN1, were identified as potential independent prognostic markers for the specific bladder cancer subtype with clinical features described.

本研究的目的是鉴定癌旁组织中差异表达的基因，并确定所选基因组的潜在预测价值。膀胱组织的基因转录组数据从UCSC Xena浏览器和NCBI GEO存储库下载，包括GTEx（基因型组织表达项目）数据、TCGA（癌症基因组图谱）数据和GEO（基因表达综合）数据。使用肿瘤-癌旁DEGs基因和癌旁-正常DEGs基因的交叉点进行差异表达基因(DEGs)分析来鉴定肿瘤-DEGs候选基因。通过Kaplan-Meier (KM) 生存分析和单变量Cox 回归筛选生存相关基因，截止标准为KM < 0.05 和cox p值< 0.05。风险模型是使用 Lasso 回归开发的。采用单因素和多因素Cox回归分析对临床数据进行分析。对高风险和低风险亚组之间的DEGs基因进行基因本体论（GO）和KEGG富集分析。我们鉴定了 6 个与生存相关的基因：EMP1、TPM1、NRP2、FGFR1、CAVIN1 和 LATS2，这些基因是在肿瘤-癌旁和癌旁-正常差异表达数据集的 DEG 分析中发现的。然后，将患者分为两类，可以根据具体的临床特征进行区分。在聚类 1 内的患者中构建了三基因风险预测模型（EMP1、FGFR1 和 CAVIN1）。该模型用于将聚类 1 患者分为高风险和低风险亚组。预后风险评分被认为是独立的预后因素。六个已确定的生存相关基因可用于膀胱癌特定亚型的分子表征。与膀胱癌的其他亚型相比，该亚型具有 T（地形）、N（淋巴结）、分期、分级和生存状态等独特的临床特征。在六个已确定的生存相关基因中，EMP1、FGFR1 和 CAVIN1 这三个基因被确定为具有描述的临床特征的特定膀胱癌亚型的潜在独立预后标志物。