摘要目的:通过TCGA数据库中膀胱尿路上皮癌(BLCA)中炎症反应相关基因(IRRGS)的表达数据和临床数据，构建BLCA分子亚型和预后分子标签，探索其在BLCA中的预后作用。方法:检索TCGA数据库中412例BLCA患者的转录组数据及临床预后数据，利用R语言中NMF包对BLCA中的IRRGS表达进行无监督聚类分析，构建基于IRRGS的BLCA分子亚型，探索不同亚型的预后情况和肿瘤微环境情况；采用LASSO-Cox回归进行BLCA预后相关甲基化标志物筛选，并构建预后分子标签，进一步分析其在BLCA中的预后作用。结果:基于BLCA转录组数据和生存数据，发现200种IRRGS中有33种基因与BLCA预后相关(均 P<0.05)，基于33种IRRGS构建BLCA分子亚型，将412例BLCA患者分为3个亚组，不同亚型具有不同预后，且不同亚型具有不同肿瘤微环境；基于33种IRRGS采用LASSO-Cox回归分析从中筛选出9种与BLCA预后相关IRRGS(DCBLD2、IL-10、IRAK2、IRF1、LDLR、PVR、RIPK2、SEMA4D、TLR2)，并联合9种IRRGS构建预后分子标签，结果显示，该分子标签可以很好地判断BLCA患者预后。 结论:基于IRRGS构建的BLCA分子亚型，不同亚型间预后不同；不同亚型间肿瘤微环境不同。基于9种IRRGS构建的BLCA预后分子标签可以很好地判断BLCA预后，提示IRRGS在BLCA中可发挥重要作用。

abstractsObjective:To construct BLCA molecular subtypes and prognostic Molecular signature through the expression data and clinical data of inflammatory response related genes (IRRGS) in bladder urothelial carcinoma (BLCA) in TCGA database, and to explore its prognostic role in BLCA.Methods:Firstly, the transcriptome data and clinical prognosis data of 412 patients with BLCA in TCGA database were downloaded. The expression of BLCA IRRGS was analyzed by unsupervised cluster analysis using NMF package in R language, and BLCA molecular subtypes based on IRRGS were constructed to explore the prognosis of different subtypes and tumor microenvironment. LASSO-Cox regression was used to screen the methylation markers related to the prognosis of BLCA, and construct the prognostic molecular tag to further explore its prognostic role in BLCA.Results:Based on BLCA transcriptome data and survival data, 33 of the 200 IRRGS were found to be related to the prognosis of BLCA (all P<0.05). BLCA molecular subtypes were constructed based on 33 IRRGS. Four hundred and twelve patients with BLCA were divided into three subgroups. Different subtypes had different prognosis, and different subtypes had different tumor microenvironment. Based on 33 IRRGS, nine IRRGS (DCBLD2, IL10, IRAK2, IRF1, LDLR, PVR, RIPK2, SEMA4D and TLR2) were selected by LASSO-Cox regression analysis. Combined with 9 IRRGS, prognostic molecular signature was constructed. Molecular signature could well judge the prognosis of BLCA patients. Conclusions:BLCA molecular subtypes are constructed based on IRRGS, and the prognosis is different among different subtypes. The tumor microenvironment is different among different subtypes. BLCA prognosis Molecular signature based on 9 IRRGS can well judge the prognosis of BLCA. These results suggest that IRRGS may play an important role in BLCA.