异常糖基化引起了生物标志物发现的极大兴趣。然而，检测能力低、聚糖结构复杂和异质性给糖蛋白检测开发带来了挑战。以触​​珠蛋白（Hp）为模型，我们开发了一个结合功能化磁性纳米颗粒和两性离子亲水相互作用液相色谱（ZIC-HILIC）的集成平台，用于高度特异性的糖肽富集，然后采用数据独立采集（DIA）策略来建立深度分析乙型肝炎病毒 (HBV, n = 5) 和肝细胞癌 (HCC, n = 5) 患者的癌症特异性 Hp 糖基化谱。 DIA 策略在血清样本中建立了最深的 Hp 糖基化景观之一（1029 个糖肽，130 个聚糖），其中包括仅在 HCC 患者中检测到的 54 种糖肽。此外，针对基于 DIA 的光谱库的单次 DIA 搜索在患者中的糖肽覆盖率优于 DDA 方法 2-3 倍。在 Hp 上的四个 N-聚糖位点（N-184、N-207、N-211、N-241）中，总聚糖类型分布显示组合岩藻糖基化-唾液酸化聚糖的检测显着增强，这是已识别的最主要的糖型在肝癌患者中。定量分析显示，48 种糖肽在 HCC 中显着富集 (p < 0.05)，其中包括 N-184 位点上的混合单唾液酸化三触角糖肽，几乎没有升高，可将 HCC 与 HBV 组区分开来（HCC/HBV 比率：2462 ± 766，p < 0.05）。总之，DIA-MS 为靶向糖蛋白组学提供了一种公正且全面的替代方案，以指导糖生物标志物的发现和验证。

Aberrant glycosylation has gained significant interest for biomarker discovery. However, low detectability, complex glycan structures, and heterogeneity present challenges in glycoprotein assay development. Using haptoglobin (Hp) as a model, we developed an integrated platform combining functionalized magnetic nanoparticles and zwitterionic hydrophilic interaction liquid chromatography (ZIC-HILIC) for highly specific glycopeptide enrichment, followed by a data-independent acquisition (DIA) strategy to establish a deep cancer-specific Hp-glycosylation profile in hepatitis B virus (HBV, n = 5) and hepatocellular carcinoma (HCC, n = 5) patients. The DIA strategy established one of the deepest Hp-glycosylation landscapes (1029 glycopeptides, 130 glycans) across serum samples, including 54 glycopeptides exclusively detected in HCC patients. Additionally, single-shot DIA searches against a DIA-based spectral library outperformed the DDA approach by 2-3-fold glycopeptide coverage across patients. Among the four N-glycan sites on Hp (N-184, N-207, N-211, N-241), the total glycan type distribution revealed significantly enhanced detection of combined fucosylated-sialylated glycans, which were the most dominant glycoforms identified in HCC patients. Quantitation analysis revealed 48 glycopeptides significantly enriched in HCC (p < 0.05), including a hybrid monosialylated triantennary glycopeptide on the N-184 site with nearly none-to-all elevation to differentiate HCC from the HBV group (HCC/HBV ratio: 2462 ± 766, p < 0.05). In summary, DIA-MS presents an unbiased and comprehensive alternative for targeted glycoproteomics to guide discovery and validation of glyco-biomarkers.