摘要
消栓通络方临床上主治脑血栓引起的精神呆滞、言语迟涩等症状,疗效显著,但其作用机制尚不明确。本文通过搜集消栓通络方中的化学成分和治疗脑卒中相关靶点,获得1251个化学成分和10个脑卒中相关靶点,采用朴素贝叶斯和递归分割等机器学习算法,基于分子指纹和分子描述符,结合分子对接方法,构建了脑卒中10个相关靶点的18个化合物-靶点相互作用预测模型。应用这些模型预测了消栓通络方的活性化学成分及其作用靶点,发现了153个潜在活性化学成分,其中22个可以与2个以上脑卒中药物靶点相互作用。在此基础上,利用网络构建专业软件,构建了化学成分-靶点网络,并通过Gene-Ontology(GO)富集分析,确证了靶点重要的生物过程,如凝血(blood coagulation)、血管生成正调控(positive regulation of angiogenesis)和离子转运正调控(positive regulation of ion transport)等。在此基础上,构建了靶点-通路网络。本研究利用机器学习、分子对接、虚拟筛选、数据挖掘及网络构建等技术方法,探索并部分揭示了消栓通络方抗脑卒中的活性物质基础及其化学成分-靶点-通路的网络作用机制,为消栓通络方的深入研究提供重要信息依据。
Xiaoshuan Tongluo formula is effective in treating mental retardation and speech astringency caused by cerebral thrombosis,but its mechanism remains unclear.In this investigation,by collecting the chemical constituents from Xiaoshuan Tongluo formula and the targets related to stroke,we obtained 1251 constituents from the formula and 10 drug targets related with stroke.We established 18 prediction models of compound-target interaction for 10 stroke-related targets,using molecular docking method and machine learning methods includes Naive Bayesian and recursive partitioning based on the input of molecular fingerprints and molecular descriptors.Using these models,we predicted the active chemical constituents from Xiaoshuan Tongluo formula and their drug targets,153 potential active constituents were discovered,22 among them could interact with at least two drug targets related with stroke.On this basis,the chemical constituent-target network was constructed using network construction software,and then the important metabolic pathways of the targets were identified by using Gene-Ontology(GO)enrichment analysis,such as blood coagulation,positive regulation of angiogenesis,positive regulation of ion transport and so on.On this basis,a target-pathway network was constructed.In conclusion,using machine learning,molecular docking,virtual screening,data mining and network construction,this study explored and partially revealed the active chemical constituents and chemical constituent-target-pathway network action mechanism of Xiaoshuan Tongluo formula against stroke,which will provide important information for its further study.
作者
郑一夫
孔令雷
贾皓
张宝月
王喆
许律捷
刘艾林
杜冠华
ZHENG Yi-fu;KONG Ling-lei;JIA Hao;ZHANG Bao-yue;WANG Zhe;XU Lü-jie;LIU Ai-liu;DU Guan-hua(Institute of Materia Medica,Chinese Academy of Medical Sciences and Peking Union Medical College,Beijing 100050,China;College of Pharmacy,Wuhan University,Wuhan 430072,China)
出处
《药学学报》
CAS
CSCD
北大核心
2020年第2期256-264,共9页
Acta Pharmaceutica Sinica
基金
国家自然科学基金资助项目(81673480)
北京市自然科学基金项目(7182113,7192134)
“十三五”重大新药创制专项(2018ZX09711001-009-009,2018ZX09711001-003-002)
协和创新工程项目(2016-I2M-3-007)
关键词
消栓通络方
脑卒中
药物靶点
机器学习
分子对接
虚拟筛选
网络药理学
Xiaoshuan Tongluo formula
stroke
drug target
machine learning
molecular docking
virtual screening
network pharmacology
