抗乳腺癌候选药物的优化模型  

Optimized model of anti-breast cancer candidate drugs

作　　者：董奕鑫 张欢欢 王昌会 陈昊[1] 李孝诚[1] DONG Yixin;ZHANG Huanhuan;WANG Changhui;CHEN Hao;LI Xiaocheng(School of Mathematical Sciences,Huaibei Normal University,Huaibei 235000,China)

机构地区：[1]淮北师范大学数学科学学院,安徽淮北235000

出　　处：《高师理科学刊》2022年第6期30-37,共8页Journal of Science of Teachers＇College and University

基　　金：安徽省自然科学研究项目(1908085MF186);安徽省高校自然科学研究重点项目(KJ2019A0589);安徽省质量工程项目(2020jyxm1670,2020jxtd)。

摘　　要：根据华为杯中国研究生数学建模竞赛D题所提供的ERα拮抗剂信息,综合运用灰色关联度分析、BP神经网络、决策树、回归模型等方法和理论,借助MATLAB,SPSS,GeoGebra等软件,构建了化合物生物活性的定量预测模型和ADMET性质分类预测模型,并在此基础上建立抗乳腺癌候选药物的优化模型,筛选出适合用于抗乳腺癌候选药物的化合物.经检验发现,模型均具有良好的性能,可将其应用于虚拟药物筛选流程,为计算机辅助药物设计与药物发现提供参考.The research data were obtained from the information of ERα antagonists provided by the D problem of Huawei Cup,a mathematical modeling competition for graduate students in China.The quantitative prediction model for the biological activity of compounds and the ADMET property classification prediction model were constructed by combining the methods and theories of gray correlation analysis,BP neural network,decision tree and regression model with the help of MATLAB,SPSS,GeoGebra.On basis of it,an optimized model of anti-breast cancer candidate drugs was established,and compounds suitable for anti-breast cancer candidate drugs were screened out.After testing,it is found that the models have good performance and can be applied to the virtual drug screening process to provide reference for computer-aided drug design and drug discovery.

关 键 词：抗乳腺癌候选药品 灰色关联度分析 BP神经网络 决策树分类预测模型 

分 类 号：O22[理学—运筹学与控制论]