基于机器学习构建乳腺癌糖基化相关基因新预后模型及药物敏感性探索

The Construction of a New Prognostic Model of Breast Cancer and the Exploration of Drug Sensitivity Based on Machine Learning for Glycosylation-Related Genes.

作者信息

Zhang Jia-Ning, Zhou Xi-Rui, Yi Zi-Lu, Tian Xin-Yu, Liu Hong

机构信息

The Second Surgical Department of Breast Cancer, Tianjin Medical University Cancer Institute & Hospital, National Clinical Research Center for Cancer, Tianjin 300060, China; Tianjin's Clinical Research Center for Cancer, Tianjin 300060, China; Key Laboratory of Breast Cancer Prevention and Therapy, Tianjin Medical University, Ministry of Education, Tianjin 300060, China.

出版信息

Clin Breast Cancer. 2025 Aug;25(6):e756-e764. doi: 10.1016/j.clbc.2025.05.004. Epub 2025 May 10.

DOI:10.1016/j.clbc.2025.05.004

PMID:40473526

Abstract

AIMS

Breast cancer has become the number 1 killer threatening women's health. In recent years, glycosylation modification has played an increasingly important role in tumor progression. The aim of this study was to explore the key genes that may be involved in glycosylation modification, establish prognostic models, and further explore their biological functions.

METHODS

Using data from TCGA and GEO databases, differentially expressed genes (DEGs) were identified. Subsequently, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were conducted to characterize the functions of the DEGs. LASSO regression analysis was performed to narrow down hub genes. Additionally, single-cell analysis, protein-protein interaction (PPI) network analysis, immune correlation analysis, drug sensitivity analysis, and molecular docking were carried out to investigate the functions of these hub genes.

RESULTS

Initially, we identified 110 differentially expressed prognostic genes, among which 89 were potentially associated with glycosylation modification. Enrichment analysis revealed their involvement in oxytocin signaling, chemical carcinogen-DNA adduct formation, and C-type lectin receptor pathways. LASSO regression (Least Absolute Shrinkage and Selection Operator) analysis further refined the selection to 24 hub genes, which exhibited specific genetic interactions. Notably, the expression levels of these genes showed significant associations with various immune cells. Drug sensitivity analysis of the hub genes highlighted methotrexate as a potential therapeutic candidate. Finally, molecular docking demonstrated strong binding affinities between the target receptors and ligands.

CONCLUSIONS

In conclusion, we screened glycosylation-related Hub genes, constructed prognostic models, explored their biological functions, and proposed new insights for diagnosing and treating breast cancer.

摘要

目的

乳腺癌已成为威胁女性健康的头号杀手。近年来，糖基化修饰在肿瘤进展中发挥着越来越重要的作用。本研究旨在探索可能参与糖基化修饰的关键基因，建立预后模型，并进一步探究其生物学功能。

方法

利用来自TCGA和GEO数据库的数据，鉴定差异表达基因（DEGs）。随后，进行基因本体论（GO）和京都基因与基因组百科全书（KEGG）通路富集分析，以表征DEGs的功能。进行LASSO回归分析以缩小核心基因范围。此外，还进行了单细胞分析、蛋白质-蛋白质相互作用（PPI）网络分析、免疫相关性分析、药物敏感性分析和分子对接，以研究这些核心基因的功能。

结果

最初，我们鉴定出110个差异表达的预后基因，其中89个可能与糖基化修饰有关。富集分析显示它们参与催产素信号传导、化学致癌物-DNA加合物形成和C型凝集素受体途径。LASSO回归（最小绝对收缩和选择算子）分析进一步将选择范围缩小至24个核心基因，这些基因表现出特定的遗传相互作用。值得注意的是，这些基因的表达水平与各种免疫细胞显示出显著关联。核心基因的药物敏感性分析突出了甲氨蝶呤作为潜在的治疗候选药物。最后，分子对接证明了靶受体与配体之间有很强的结合亲和力。