一种改进的网络和化学信息学驱动的化合物作用机制识别方法
详细信息 查看官网全文
- 英文论文题名：Identification of chemical mechanism-of-action through an improved network and chemoinformatics-driven approach
- 论文作者：吴曾睿 ; 吴荡 ; 卞汉平 ; 李杰 ; 黄瑾 ; 唐赟
- 英文论文作者：Zengrui Wu ; Dang Wu ; Hanping Bian ; Jie Li ; Jin Huang ; Yun Tang ; Shanghai Key Laboratory of New Drug Design ; School of Pharmacy ; East China University of Science and Technology
- 年：2016
- 作者机构：华东理工大学药学院上海市新药设计重点实验室;
- 论文关键词：药物-靶标相互作用 ; 药物重定位 ; 基于网络推理 ; 作用机制 ; 雌激素受体
- 会议召开时间：2016-07-01
- 会议录名称：中国化学会第30届学术年会摘要集-第二十五分会：化学信息学与化学计量学
- 语种：中文
- 分类号：TQ460.1
- 学会代码：ZGHY
- 会议名称：中国化学会第30届学术年会-第二十五分会 ; 化学信息学与化学计量学
- 会议地点：中国辽宁大连
- 主办单位：中国化学会
- 学会名称：中国化学会
- 页数：1
- 文件大小：213k
- 原文格式：D
- 会议级别：全国
摘要
阐明化合物的作用机制,有助于新型治疗用途的发现(例如药物重定位)以及药物副作用的评估。在系统药理学的框架下,开发用于评价化合物作用机制的新型计算方法,可以快速、廉价地加速药物发现和开发过程。本研究中,我们开发了一种改进的网络和化学信息学驱动的靶标预测方法,名为平衡的基于子结构—药物—靶标网络推理(bSDTNBI)方法,可以为老药、临床失败药物或全新化学实体,预测其潜在的作用机制。这一方法基于我们前期开发的SDTNBI方法,在基于网络的资源扩散过程中引入了三个参数,分别用于调整网络中不同类型节点的初始资源分配、不同类型边的权重设置、枢纽节点的影响,提升了方法的预测准确率。bSDTNBI方法构建的预测模型,在十折交叉验证和留一交叉验证中均表现良好。此外,bSDTNBI方法预测出的27个化合物,在体外实验测试中显示出了与雌激素受体α的结合活性(IC_(50)或EC_(50)≤10μM)。总之,bSDTNBI方法可以为药物发现和开发中的化合物作用机制评价提供有力工具。
Deciphering chemical mechanism-of-action(MoA) enables the discovery of novel therapeutics(e.g.drug repositioning) and evaluation of drug side effects.Development of novel computational approaches for chemical MoA assessment under systems pharmacology framework would accelerate drug discovery and development with high efficiency and low cost.In this study,we proposed an improved network and chemoinformatics-driven approach,namely balanced substructure-drug-target network-based inference(bSDTNBI),to predict potential MoA for old drugs,clinical failed drugs,and new chemical entities.High performance was yielded for the models built via bSDTNBI in both 10-fold cross validation and leave-one-out cross validation.In a case study,27 predicted compounds were experimentally validated to have high binding affinities on estrogen receptor a with IC_(50) or EC_(50)values less than 10μM.In summary,bSDTNBI would provide a powerful tool for the chemical MoA assessment during drug discovery and development.
Deciphering chemical mechanism-of-action(MoA) enables the discovery of novel therapeutics(e.g.drug repositioning) and evaluation of drug side effects.Development of novel computational approaches for chemical MoA assessment under systems pharmacology framework would accelerate drug discovery and development with high efficiency and low cost.In this study,we proposed an improved network and chemoinformatics-driven approach,namely balanced substructure-drug-target network-based inference(bSDTNBI),to predict potential MoA for old drugs,clinical failed drugs,and new chemical entities.High performance was yielded for the models built via bSDTNBI in both 10-fold cross validation and leave-one-out cross validation.In a case study,27 predicted compounds were experimentally validated to have high binding affinities on estrogen receptor a with IC_(50) or EC_(50)values less than 10μM.In summary,bSDTNBI would provide a powerful tool for the chemical MoA assessment during drug discovery and development.
引文
[1]Wu,Z.;Cheng,F.;Li,J.;et al.Brief Bioinform.,2016,doi:10.1093/bib/bbw012.
[2]Cheng,F.;Zhou,Y.;Li,W.;et al.PLoS One,2012,7(7):e41064.
[3]Cheng,F.;Liu,C;Jiang,J.;et al.PLoS Comput.Biol.,2012,8(5):el002503.
[2]Cheng,F.;Zhou,Y.;Li,W.;et al.PLoS One,2012,7(7):e41064.
[3]Cheng,F.;Liu,C;Jiang,J.;et al.PLoS Comput.Biol.,2012,8(5):el002503.