基于数据挖掘及网络药理学分析新安医家治疗消渴的用药规律及作用机制  

Medication rules and mechanism of Xin’an physicians in the treatment of diabetes mellitus based on data mining and network pharmacology

作　　者：朱晋如 吴晴 胡可心 王梦凡 方朝晖 ZHU Jinru;WU Qing;HU Kexin;WANG Mengfan;FANG Zhaohui(First Clinical College of Anhui University of Traditional Chinese Medicine,Hefei 230031,China;Center for Xin’an Medicine and Modernization of Traditional Chinese Medicine of Institute of Health and Medicine,Hefei 230012,China;Department of Endocrinology,First Affiliated Hospital of Anhui University of Traditional Chinese Medicine,Hefei 230031,China;Key Laboratory of the Ministry of Education of the Xin’an Medicine,Anhui University of Chinese Medicine,Hefei 230031,China;Famous Fang Zhaohui Studio of Chinese Medicine in Anhui Province,Hefei 230031,China)

机构地区：[1]安徽中医药大学第一临床医学院,合肥230031 [2]合肥综合性国家科学中心大健康研究院新安医学与中医药现代化研究所,合肥230012 [3]安徽中医药大学第一附属医院内分泌科,合肥230031 [4]安徽中医药大学新安医学教育部重点实验室,合肥230031 [5]方朝晖安徽省名中医工作室,合肥230031

出　　处：《药物流行病学杂志》2025年第3期260-272,共13页Chinese Journal of Pharmacoepidemiology

基　　金：国家自然科学基金面上项目(82174153);安徽高校协同创新项目(GXXT-2020-025);合肥综合性国家科学中心大健康研究院“揭榜挂帅”项目(2023CXMMTCM003)。

摘　　要：目的挖掘新安医家治疗消渴的用药规律,并分析核心处方中药物的作用机制。方法收集新安医学著作中治疗消渴的处方,使用SPSS Modeler 18.0软件进行配伍关联规则以及复杂网络分析,获得核心药物处方,并利用SPSS Statistics 25.0软件进行高频中药聚类分析。通过BATMAN-TCM数据库提取核心药物的有效活性成分及作用靶点;GeneCards数据库检索糖尿病基因靶点,获取交集靶点后,导入STRING在线平台构建蛋白互作网络,使用Cytoscape 3.9.1软件筛选核心基因;并导入DAVID数据库进行基因本体(GO)功能及京都基因和基因组百科全书(KEGG)通路富集分析。结果共收录有效处方135条,涉及中药184味,药性以寒、平为主;药味以甘、苦、辛为主;主归肺、脾、肾、胃、心经。关联规则挖掘常用药对14对。聚类分析将前20味中药聚为5个聚类组。复杂网络可视化分析形成了由炙甘草、人参、知母、麦冬、茯苓组成的核心处方。筛选出核心处方药物的有效活性成分164个,作用靶点1498个,糖尿病基因靶点1995个,交集靶点404个,提取核心靶点10个,GO功能富集分析共得到涵盖生物过程(BP)1363条,涉及细胞组分(CC)129条,包含分子功能(MF)264条;涉及KEGG信号通路206条。结论新安医家治疗消渴的遣方用药体现了新安医家“标本兼治”的学术思想以及“善用甘寒、辛开苦降、培土生金、固本培元”的治疗思路,为现代中医临床提供了新的糖尿病诊疗思路和启示。Objective To explore the medication rules and mechanism of Xin'an physicians in the treatment of diabetes mellitus.Methods The prescriptions for the treatment of diabetes mellitus in Xin'an medical writings were collected.The core drug prescriptions were obtained by using IBM SPSS Modeler 18.0 for dispensing association rules as well as complex network analysis,and the cluster analysis of high-frequency traditional Chinese medicines was carried out by using IBM SPSS Statistics 25.0.The active ingredients and targets of the core drugs were extracted from the BATMAN-TCM database.The GeneCards database was used to search for diabetes gene targets,and after obtaining the intersecting targets,the STRING online platform was imported to construct the protein interactions network,and screened the core genes by using the Cytoscape 3.9.1 software,and imported into the DAVID database for GO function and KEGG pathway enrichment analysis.Results A total of 135 valid prescriptions were included,involving 184 flavors of traditional Chinese medicines,with cold and mild properties.Sweet,bitter and pungent flavors,and the main attributes of lung,spleen,kidney,stomach and heart meridians.The association rules mined 14 pairs of commonly used medicine pairs.Cluster analysis clustered the top 20 traditional Chinese medicines into five cluster groups.Complex network visualization analysis formed a core prescription consisting of honey-fried licorice root,ginseng,rhizoma anemarrhenae,ophiopogon japonicus,poria cocos.The core prescription drugs were screened for 164 effective active ingredients,1,498 action targets,1,995 diabetes gene targets,404 intersecting targets,10 core targets were extracted,and a total of 1,363 items covering BP were obtained from GO functional enrichment analysis,129 items involving CC,264 items containing MF,and 206 items of KEGG signaling pathway.Conclusion The prescriptions and medicines used by Xin'an doctors in the treatment of diabetes mellitus reflect the academic idea of"treating both the symptoms and the

关 键 词：新安医学 消渴 数据挖掘 网络药理学 用药规律 作用机制 

分 类 号：R289[医药卫生—方剂学]