通过索引药品说明书信息挖掘药物新用途  

Drug Label Indexing and Mining for Drug Repurposing

作　　者：杨立[1] 王可鉴[1] 贺林[1] 

机构地区：[1]上海交通大学Bio-X研究院,上海200030

出　　处：《复旦学报（自然科学版）》2013年第6期755-761,F0003,共8页Journal of Fudan University：Natural Science

摘　　要：通过基于文本的数据挖掘技术,对1 190个药物的12 244个不同的电子药物说明书进行收集整理,将药物说明书中描述药物信息的自然语言索引成一体化医学语言系统下的编码词条.在此基础上研究了不良反应与适应症之间的关联,并发现了多组显著的内在关联.其中,眼球突出症作为不良反应与适应症骨关节炎的关联,以及不良反应碱中毒与适应症过敏性鼻炎的关联都得到了实验的验证.同时发现药物乙酰唑胺潜在的治疗抑郁症价值,以及药物咪达唑仑可能改善癫痫的作用,对新药物的发现具有潜在应用价值.By text-based data-mining method, the project collected and organized 12 244 electronic drug-label documents of 1 190 drugs, and indexed the medicine information of those drug labels as coded entries in Unified Medical Language System（UMLS）. On this basis, we studied the correlation between side-effect and indication and revealed many significantly correlated ＂side effeet-indication＂pairs. Among these pairs, the correlations of the ＂exophthalmos-osteoarthritis＂ and ＂alkalosis-allergic rhinitis＂ were supported by medical trials. Besides, the project also discovered that Acetazolamide could be used to treat depressive disorder, and Midazolam have potential use for seizures. Which may have potential applicative value for the discovery of new drug.

关 键 词：药物说明书 老药新用 一体化医学语言系统 

分 类 号：R857.3[医药卫生—航空、航天与航海医学]