计算机工程与应用 ›› 2017, Vol. 53 ›› Issue (9): 11-16.DOI: 10.3778/j.issn.1002-8331.1612-0111

• 热点与综述 • 上一篇    下一篇

机器学习预测ACEI的比较与分析

胡明伟1,丁彦蕊2,3   

  1. 1.江南大学 物联网工程学院,江苏 无锡 214122
    2.江南大学 数字媒体学院,江苏 无锡 214122
    3.江南大学 工业生物技术教育部重点实验室,江苏 无锡 214122
  • 出版日期:2017-05-01 发布日期:2017-05-15

Comparison and analysis of machine learning prediction of ACEI

HU Mingwei1, DING Yanrui2,3   

  1. 1.School of Internet of Things Engineering, Jiangnan University, Wuxi, Jiangsu 214122, China
    2.School of Digital Media, Jiangnan University, Wuxi, Jiangsu 214122, China
    3.Key Laboratory of Industrial Biotechnology, Jiangnan University, Wuxi, Jiangsu 214122, China
  • Online:2017-05-01 Published:2017-05-15

摘要: 血管紧张素转换酶抑制剂(ACEI)对高血压的治疗具有重要意义。基于从结构复杂的化合物数据库中构建的候选小分子数据集,采用分子对接技术从数据集中筛选出样本构建分类模型。分别采用支持向量机、[K]近邻、决策树、随机森林和贝叶斯方法建立血管紧张素转换酶潜在抑制剂和非抑制剂的分类模型。经结果对比,支持向量机相比于其他方法有更高的预测率,其中模型总体预测率和相关系数分别为82.4%和0.653。研究表明,支持向量机方法对于虚拟筛选血管紧张素转换酶抑制剂具有良好的效果。

关键词: 血管紧张素转换酶抑制剂(ACEI), 分子对接, 机器学习, 支持向量机

Abstract: Angiotensin Converting Enzyme Inhibitor(ACEI) plays an important role in the treatment of hypertension. Candidate small molecular data sets are constructed from the database of complex compounds and the sample sets obtained from the data set using molecular docking techniques are used to construct the classification model. The classification model of angiotensin converting enzyme inhibitors and non inhibitors is established by using support vector machine, [K] nearest neighbor, decision tree, random forest and Naive Bayes method, respectively. The support vector machine has higher prediction rate compared with other methods and the overall prediction and correlation coefficients of the model are 82. 4% and 0. 653, respectively. The results show that the support vector machine method has a good effect on the virtual screening of angiotensin converting enzyme inhibitors.

Key words: Angiotensin Converting Enzyme Inhibitor(ACEI), molecular docking, machine learning, support vector machine