组织工程与重建外科杂志

组织工程与重建外科杂志 ›› 2025, Vol. 21 ›› Issue (3): 238-.

• • 上一篇    下一篇

通过整合生物信息学分析与机器学习揭示糖尿病足溃疡缺氧相关生物标志物

  

  • 出版日期:2025-06-02 发布日期:2025-07-01

 Unveiling hypoxia-related biomarkers for diabetic foot ulcers through integrated bioinformatics analysis and machine learning

  • Online:2025-06-02 Published:2025-07-01

PDF (PC)

3

摘要： 目的 糖尿病足溃疡（Diabetic foot ulcer，DFU）是糖尿病的严重并发症，缺氧微环境在其发生和愈合延迟中起 关键作用，但其分子机制尚不明确。本研究拟通过生物信息学方法，系统解析 DFU中缺氧相关基因的调控网络，筛选 关键生物标志物，为靶向治疗提供依据。方法 整合 GEO和 MSigDB缺氧数据集进行差异表达分析、加权基因共表达 网络分析（WGCNA）、GO/KEGG 功能富集，并通过 Lasso、SVM-RFE 和随机森林三种机器学习算法筛选枢纽基因，验证 其诊断效能。结果 本研究共鉴定出152个差异表达基因（DEGs），包括14个缺氧相关差异表达基因（HRDEGs），富集 分析显示 HRDEGs参与葡萄糖代谢、脂质代谢、免疫细胞调节等生物学过程。机器学习进一步筛选出枢纽基因 BGN。 BGN在DFU组中呈现低表达，训练集和验证集的ROC曲线下面积（AUC）分别为0.833和0.931，显示高诊断价值。单基因 GSEA表明，BGN通过调控组织修复、炎症反应和细胞外基质交互参与DFU病理进程。结论 BGN是DFU缺氧微环境的 关键生物标志物，可能成为早期诊断和靶向治疗的潜在分子靶点。本研究为DFU的机制解析和临床干预提供了新方向。

关键词: 糖尿病足溃疡, &emsp, 缺氧, &emsp, 生物信息学

Abstract:   Objective  Diabetic foot ulcer （DFU） is a severe complication in diabetic patients， where the hypoxic microenvironment plays a critical role in its pathogenesis and delayed healing， though the underlying molecular mechanisms remain unclear. To systematically analyze the regulatory network of hypoxia-related genes in DFU using bioinformatics approaches， identify key biomarkers， and provide insights for targeted therapies. Methods Integrated datasets from GEO and MSigDB hypoxia-related gene sets were utilized. Differential expression analysis（ limma， DESeq2）， weighted gene coexpression network analysis （WGCNA）， and GO/KEGG functional enrichment were performed. Hub genes were screened using three machine learning algorithms（ Lasso， SVM-RFE， and random forest）， and their diagnostic efficacy was validated. Results  A total of 152 differentially expressed genes （DEGs） were identified， including 14 hypoxia-related DEGs （HRDEGs）. Enrichment analysis revealed HRDEGs involvement in glucose metabolism， lipid metabolism， and immune cell regulation. Machine learning further pinpointed the hub gene BGN. BGN exhibited significantly lower expression in DFU groups， with area under the ROC curve （AUC） values of 0.833 （training set） and 0.931 （validation set）， indicating high diagnostic accuracy. Single-gene GSEA demonstrated that BGN participates in DFU pathology by regulating tissue repair， inflammatory responses， and extracellular matrix interactions. Conclusion BGN is a key biomarker in the hypoxic microenvironment of DFU and may serve as a potential molecular target for early diagnosis and targeted therapy. This study provides new directions for understanding DFU mechanisms and clinical interventions.

Key words: &emsp, Diabetic foot ulcer, &emsp, Hypoxia, &emsp, Bioinformatics