基于生物信息学、网络药理学和分子对接探究甘草附子汤治疗强直性脊柱炎的机制  

Based on bioinformatics,network pharmacology and molecular docking,the mechanism of Gancao Fuzi decoction in the treatment of ankylosing spondylitis were explored

作　　者：易振 刘金萍 吴婷 殷海东 贺宪[2] YI Zhen;LIU Jinping;WU Ting;YIN Haidong;HE Xian(The Graduate School of Guangzhou University of Chinese Medicine,Guangzhou,Guangdong 510006,China;Department of Orthopedics,Panyu District Hospital of Traditional Chinese Medicine,Guangzhou,Guangdong 511400,China)

机构地区：[1]广州中医药大学研究生院,广东广州510006 [2]广州市番禺区中医院骨科,广东广州511400

出　　处：《安徽医药》2024年第12期2359-2367,I0005,共10页Anhui Medical and Pharmaceutical Journal

摘　　要：目的基于生物信息学、网络药理学和分子对接探究甘草附子汤治疗强直性脊柱炎(AS)的机制。方法从GEO数据库检索2023年7月以前与AS相关的基因表达谱数据获得GSE25101、GSE73754基因芯片,在R语言中对2个芯片进行整合,利用limma包筛选差异表达基因(DEGs),同时在GenegCards、Msigdb、IMMPort 3个数据库中下载免疫相关基因并取交集。基于DEGs与3个数据库免疫相关基因的交集取交集,再筛选出相关免疫-差异表达基因(IDEG)的相关矩阵;通过构建风险模型判断IDEG的诊断效果,选择LASSO回归模型识别AS预测的特征基因,并绘制校准曲线和决策曲线,Cox比例风险回归模型分析结果使用森林图展现;采用受试者操作特征曲线(ROC曲线)验证风险模型和特征基因的表达;将DEGs和IDEG导入String数据库得到蛋白质-蛋白质相互作用(PPI),使用Cytoscape软件进行分析并进行可视化;利用clusterProfiler包对IDEG进行基因本体(GO)和京都基因和基因组数据库(KEGG)富集分析,筛选出特征基因。然后,利用CIBERSORT反卷积法分析16个基因与22种免疫细胞的相关性;基于TCMSP、Uniprot数据库中搜索甘草附子汤的中药活性成分及其药物靶点,利用DisGeNET、Gene‑Cards数据库检索疾病靶点,取中药靶点及疾病靶点交集后导入String数据库,使用Cytoscape软件构建疾病基因与药物共有靶点及靶点蛋白互作网络,最终,使用Autodock、Pymol进行分子对接并可视化。结果GSE25101和GSE73754并集后共得到190个DEGs,筛选出16个IDEG:CXCL8、HLA-DQA1、MMP9、TGFBR3、S100A12、PRF1、KIR2DL4、CARD11、HLA-C、ADM、ADRB2、CD81、IL2RB、APOBEC3A、KIR2DL3、CXCR1;在风险模型中,使用LASSO算法从16个基因中识别出9个最具影响力的特征基因系数,校准曲线显示,线图模型预测与理想模型预测几乎相同,并且决策曲线分析或复合遗传模型中的单一预测风险评分优于随机模型中的单一预测风险评分,9�Objective To explore the mechanism of Gancao Fuzi decoction in the treatment of ankylosing spondylitis(AS)on the basis of bioinformatics,network pharmacology and molecular docking.Methods The GSE25101 and GSE73754 gene chips were obtained from GEO database by searching gene expression profiling data related to AS before July 2023,the 2 chips were integrated in R lan-guage,and differentially expressed genes(DEGs)were screened by using the limma package,and meanwhile,immune-related genes were downloaded from 3 databases,namely GenegCards,Msigdb and IMMPort,and the intersection was taken.Based on the intersec-tion of DEGs and immune-related genes in three databases,the correlation matrix of related immune-differentially expressed genes(IDEG)was screened.The diagnostic effect of IDEG was judged by constructing a risk model.The LASSO regression model was select-ed to identify the characteristic genes predicted by AS,and the calibration curve and decision curve were drawn.The Cox regression analysis results were shown by forest map.The receiver operating characteristic(ROC)curve was used to verify the risk model and the expression of characteristic genes.DEGs and IDEG were imported into String database to obtain protein-protein interaction(PPI),which was analyzed and visualized by Cytoscape software.The clusterProfiler package was used to perform gene ontology(GO)and Kyo-to encyclopedia of genes and genomes(KEGG)enrichment analysis on IDEG to screen out characteristic genes.Then,the correlation between 16 genes and 22 immune cells was analyzed by CIBERSORT deconvolution method.Based on TCMSP and Uniprot database,the active components and drug targets of Gancao Fuzi decoction were searched.The disease targets were searched in DisGeNET and GeneCards databases.The intersection of traditional Chinese medicine targets and disease targets was imported into String database.Cytoscape software was used to construct the common targets and target protein interaction network of disease genes and drugs.Finally,Autodock and Pymol we

关 键 词：脊柱炎 强直性 甘草 附子 生物信息学 白细胞介素-10 芳基烃受体 免疫相关基因 网络药理学 分子对接 

分 类 号：R285[医药卫生—中药学]