膀胱癌（BCa）是泌尿系统的主要恶性肿瘤。在此，初步建立了全面的尿液蛋白质组学特征，用于膀胱癌的无创诊断和复发监测。收集279例（63例原发BCa、87例非肿瘤对照（NT）、73例复发BCa（BCR）和56例非复发BCa（BCNR））来筛查尿蛋白生物标志物。分别通过 DDA 和两个发现组中所有理论质谱 (SWATH-MS) 分析的连续窗口采集对 4761 和 3668 种蛋白质进行定性和定量。通过多反应监测（MRM）在两个独立的组合组中验证上调的蛋白质。使用多支持向量机递归特征消除 (mSVM-RFE) 算法，包含 13 种蛋白质的模型在 BCa 和 NT 之间表现出良好的性能，AUC 为 0.821 (95% CI: 0.675-0.967)，灵敏度为 90.9% (95诊断测试集中的 % CI：72.7-100%）和 73.3% 特异性（95% CI：53.3-93.3%）。同时，11 标记物分类器显着区分 BCR 和 BCNR，敏感性为 75.0%（95% CI：50.0-100%），特异性为 81.8%（95% CI：54.5-100%），AUC 为 0.784（95% CI） ：0.609-0.959）在复发监测测试队列中。值得注意的是，新报道了 24 个标记的 6 个蛋白（SPR、AK1、CD2AP、ADGRF1、GMPS 和 C8A）。本文揭示了 BCa 的新型尿蛋白生物标志物，并为膀胱癌的发病机制提供了新的理论见解（数据标识符 PXD044896）。

Bladder cancer (BCa) is the predominant malignancy of the urinary system. Herein, a comprehensive urine proteomic feature was initially established for the noninvasive diagnosis and recurrence monitoring of bladder cancer. 279 cases (63 primary BCa, 87 nontumor controls (NT), 73 relapsed BCa (BCR), and 56 nonrelapsed BCa (BCNR)) were collected to screen urinary protein biomarkers. 4761 and 3668 proteins were qualified and quantified by DDA and sequential window acquisition of all theoretical mass spectra (SWATH-MS) analysis in two discovery sets, respectively. Upregulated proteins were validated by multiple reaction monitoring (MRM) in two independent combined sets. Using the multi-support vector machine-recursive feature elimination (mSVM-RFE) algorithm, a model comprising 13 proteins exhibited good performance between BCa and NT with an AUC of 0.821 (95% CI: 0.675-0.967), 90.9% sensitivity (95% CI: 72.7-100%), and 73.3% specificity (95% CI: 53.3-93.3%) in the diagnosis test set. Meanwhile, an 11-marker classifier significantly distinguished BCR from BCNR with 75.0% sensitivity (95% CI: 50.0-100%), 81.8% specificity (95% CI: 54.5-100%), and an AUC of 0.784 (95% CI: 0.609-0.959) in the test cohort for relapse surveillance. Notably, six proteins (SPR, AK1, CD2AP, ADGRF1, GMPS, and C8A) of 24 markers were newly reported. This paper reveals novel urinary protein biomarkers for BCa and offers new theoretical insights into the pathogenesis of bladder cancer (data identifier PXD044896).