乳腺癌是全球最常见的癌症，其病死率在所有癌症中排名第五，并具有高转移风险。然而，乳腺癌患者临床结局分层的潜在生物标志物和分子机制仍需要研究。因此，我们的目标是寻找乳腺癌患者的新的预后生物标志物和治疗靶点。我们使用无监督分层聚类方法对公开的乳腺癌转录组学数据集进行全面研究，其中包含185个糖原基因，并结合临床病理学和生存信息。利用Limma软件包发现了基于糖原基因的亚型分类特征，并通过GSVA分析确定其与四个已知的分子特征的相关性。我们进行了实验证实，并通过定量RT-PCR、免疫印迹、透明质酸酶法实验和凝集素免疫荧光染色等方法，对B3GNT7的生物功能进行了表征。我们发现了一个由23个糖原基因组成的分类标志。在这23个糖原基因中，B3GNTs与乳腺癌患者（n=2655）中的ER-/Her2-亚型显著正相关。在公开数据集中，过表达的B3GNT7与乳腺癌患者的不良预后相关。B3GNT7的消耗抑制了MDA-MB-231和HCC1937乳腺癌细胞的增殖、迁移和侵袭，并降低了全球性富含糖基化。因此，我们发现了一种独特的23基因标志，用于乳腺癌患者的糖原基因类型分类。在这些基因中，B3GNT7显示出作为不利预后和治疗靶点的潜力生物标志物。© 2023. The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.

Breast cancer is the most common cancer worldwide, with the fifth highest mortality rate among all cancers and high risk of metastasis. However, potential biomarkers and molecular mechanisms underlying the stratification of breast cancer in terms of clinical outcomes remain to be investigated. Therefore, we aimed to find a novel prognostic biomarker and therapeutic target for breast cancer patients.Unsupervised hierarchical clustering was used to perform comprehensive transcriptomic study of total 185 glycogenes in public datasets of breast cancer with clinicopathological and survival information. A glycogene-based signature for subtype classification was discovered using Limma packages, and relevance to four known molecular features was identified by GSVA. Experimental verification was performed and biological functions of B3GNT7 were characterized by quantitative RT-PCR, western blot, transwell assays, and lectin immunofluorescence staining in breast cancer cells.A 23-glycogene signature was identified for the classification of breast cancer. Among the 23 glycogenes, B3GNTs showed significantly positive associations with ER-/Her2- subtype in breast cancer patients (n = 2655). Overexpressed B3GNT7 were correlated with poor prognosis in breast cancer patients based on public datasets. B3GNT7 depletion inhibited cell proliferation, migration, and invasion, and decreased global fucosylation in MDA-MB-231 and HCC1937 breast cancer cells.Herein, we discovered a unique 23-gene signature for breast cancer patient glycogene-type classification. Among these genes, B3GNT7 was shown to be a potential biomarker for unfavorable outcomes and therapeutic target of breast cancer.© 2023. The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.