刘文斌^{1, 2,,},
陈杰¹,
方刚²,
石晓龙²,
许鹏^{2, 3}
1.温州大学计算机与人工智能学院 温州 325035
2.广州大学计算科技研究院 广州 510006
3.黔南民族师范学院计算机与信息学院 都匀 558000
基金项目:国家重点研发计划(2019YFA0706402)，国家自然科学基金(61572367, 61573017, 61972107, 61972109)

详细信息

作者简介:刘文斌：男，1969年生，教授，研究方向为生物信息学
陈杰：男，1994年生，硕士生，研究方向为生物信息学
方刚：男，1969年生，教授，研究方向为生物信息学
石晓龙：男，1975年生，教授，研究方向为生物信息学
许鹏：男，1986年生，博士后，研究方向为生物信息学

通讯作者:刘文斌　wbliu6910@126.com

中图分类号:TP301

计量

文章访问数:1608
HTML全文浏览量:543
PDF下载量:63
被引次数:0

出版历程

收稿日期:2019-11-01
修回日期:2020-01-15
网络出版日期:2020-02-18
刊出日期:2020-06-22

Prediction of Drug Synergy and Antagonism Based on Drug-Drug Interaction Network

Wenbin LIU^{1, 2,,},
Jie CHEN¹,
Gang FANG²,
Xiaolong SHI²,
Peng XU^{2, 3}
1. College of Computer Science and Artificial Intelligence, Wenzhou University, Wenzhou 325035, China
2. Institute of Computing Science and Technology, Guangzhou University, Guangzhou 510006, China
3. School of Computer Science and Information Technology, Qiannan Normal University for Nationalities, Duyun 558000, China
Funds:The National Key R&D Program of China (2019YFA0706402), The National Natural Science Foundation of China (61572367, 61573017, 61972107, 61972109)


摘要
摘要:药物的协同与拮抗关系预测，有助于药物的使用安全及组合用药的发展。该文从药物互作网络(DDINet)出发，基于网络拓扑结构构造分类特征，提出一种预测药物协同和拮抗关系的方法。从特征选择结果可知，根据药物与其公共邻居节点关系构造的特征表现出了明显的正负样本分布差距，能有效地反映出药物的协同或拮抗关系。在使用不同特征分类器的分类结果中，最优AUC和分类精度值分别达到了0.9687和0.9187。而在协同与拮抗关系预测结果中，其预测精度值达到了0.45和0.75以上。这说明基于网络拓扑结构的方法能有效对药物协同和拮抗关系进行分类和预测。与传统基于药物功能、结构、靶基因等相似性特征的方法相比，该方法计算简单高效，将会有效促进组合用药的发展。
关键词:药物相互作用预测/
网络拓扑结构/
药物协同/
药物拮抗
Abstract:Accurately predicting the synergistic and antagonistic relationship of drugs is helpful to the safety of drug use and the development of drug combination. A method for predicting drug synergy and antagonistic is proposed, which based on the Drug-Drug Interaction Network (DDINet) and its topological structure. From the result of feature selection, it can be seen that the feature constructed based on the interaction between the drug and its common neighbor node shows an obvious difference in the distribution of positive and negative samples, which can effectively reflect the drug synergy or antagonism. In the classification results using different feature classifiers, the optimal Area Under the Curve (AUC) and classification accuracy value reache 0.9687 and 0.9187 respectively. In the prediction results of synergy and antagonism, the prediction accuracy also reache above 0.45 and 0.75. This shows that the method based on network topology can effectively classify and predict the synergistic and antagonistic effects of drugs. Compared with the traditional methods based on similarity features of drug function, structure, target gene, etc, this method is simple and efficient to calculate, and can effectively promote the development of combination drugs.
Key words:Prediction Drug-Drug Interactions(DDIs)/
Network topology/
Synergy/
Antagonism



PDF全文下载地址:

https://jeit.ac.cn/article/exportPdf?id=fc30c742-e51a-45e2-801f-973f74000fc3