| 摘要: |
| 目的：通过网络药理学、机器学习法与孟德尔随机化，探究参芪降糖胶囊（SJC）治疗糖尿病肾病（DN）的潜在作用机制。方法：利用多个数据平台筛选SJC治疗DN的主要活性成分，对药物与疾病的靶标基因取交集后采用Cytoscape软件构建活性成分-靶点网络图，导入String数据库构建蛋白质-蛋白质相互作用（PPI）网络，进行拓扑分析寻找核心基因。应用GEO数据库对核心基因进行统计性分析，寻找核心基因中的差异性基因（DEG），并构建机器学习模型，得出分值前5位的基因，进行两样本孟德尔随机化（MR）分析，验证相关靶点与DN的因果关系。结果：SJC治疗DN的主要活性成分为槲皮素、山柰酚等；药物与疾病的交集靶点有89个，核心靶点包括白细胞介素-10（IL-10）、肿瘤坏死因子（TNF）、表皮生长因子（EGF）等，进行统计学分析后得出13个DEG，构建机器学习模型得出分值前5位的基因为EGF、趋化因子配体2（CCL2）、基质金属蛋白酶2（MMP2）、白细胞介素-6（IL-6）、细胞间黏液分子1（ICAM1）。MR分析研究得出IL-6与DN的发生存在因果关系。 结论：本研究利用多种方法揭示了SJC通过多靶点作用于DN以发挥药效，并验证了基因与DN的因果关系，可为临床合理用药提供参考。 |
| 关键词: 糖尿病肾病 参芪降糖胶囊 孟德尔随机化 机器学习 网络药理学 |
| DOI： |
|
|
| Mechanism of action of Shenqi Jiangtang capsules in treatment of diabetic nephropathy：A study based on network pharmacology，machine learning，and Mendelian randomization |
| CHEN Nianqing,LEI Weiyi,SHI Yifan |
| (The First School of Clinical Medicine of Hunan University of Chinese Medicine，Changsha 410007，Hunan，China) |
| Abstract: |
| Objective：To investigate the potential mechanism of action of Shenqi Jiangtang capsules (SJC) in the treatment of diabetic nephropathy (DN) based on network pharmacology，machine learning，and Mendelian randomization.Methods：Several databases were used to obtain the main active components of SJC in the treatment of DN，and after the target genes of the drug and the disease were intersected，Cytoscape software was used to construct an active component-target network.String database was imported to construct a protein-protein interaction network，and the topology analysis was performed to search for core genes.GEO database was used to perform a statistical analysis of core genes and obtain the differentially expressed genes (DEGs)，and machine learning models were constructed.The top 5 genes in terms of the score were included in the two-sample Mendelian randomization analysis，and the causal relationship between related targets and DN was validated.Results：The main active components of SJC in the treatment of DN included quercetin and kaempferol.There were 89 intersecting targets between the drug and the disease，and the core targets included interleukin-10，tumor necrosis factor，and epidermal growth factor (EGF).A total of 13 DEGs were obtained by the statistical analysis，and the machine learning models showed that the top 5 genes in terms of the score were EGF，CCL2，MMP2，IL-6，and ICAM1.The Mendelian randomization analysis showed the presence of the causal relationship between IL-6 and DN.Conclusion：By using various methods，this study shows that SJC exerts a therapeutic effect on DN through multiple targets，and it also validates the causal relationship between IL-6 and DN，which provides a reference for rational drug use in clinical practice. |
| Key words: diabetic nephropathy Shenqi Jiangtang capsules Mendelian randomization machine learning network pharmacology |
