用支持向量机预测HIV-1整合酶抑制剂活性  

Support Vector Machine Applied to Predicting the Activity of HIV-1 IN Inhibitors

作　　者：刘昕[1] 谭建军[1] 陈慰祖[1] 刘斌[1] 李杉[1] 王存新[1] 

机构地区：[1]北京工业大学生命科学与生物工程学院,北京100124

出　　处：《北京工业大学学报》2013年第4期634-640,共7页Journal of Beijing University of Technology

基　　金：国家自然科学基金资助项目(21173014);科技部国际合作项目(2010DFA31710)

摘　　要：HIV整合酶可催化病毒复制周期中的整合过程,即将HIV反转录产物cDNA整合入宿主基因组,它是病毒复制过程中不可缺少的酶,也是抗HIV药物设计的重要靶点.构建嘧啶酮类(pyrimidones)HIV-1整合酶抑制剂定量构效关系模型,有助于进一步了解影响抑制剂活性的结构因素.本文应用CoMFA软件计算了68个化合物的拓扑、分子极化、亲水性等结构参数,用所选的结构参数作为支持向量机(support vector machine,SVM)的输入,建立起非线性的支持向量机回归模型.研究表明:支持向量机算法与分子结构参数的有机集成,可为HIV整合酶抑制剂的结构与活性数据建立起预测模型,为抗HIV药物设计提供生物学信息.HIV integrase （IN） catalyzes the integration process in the viral life cycle. IN helps the viral reverse transcription product cDNA integrating into the host chromosome. As an indispensable enzyme, IN is also an important target for designing and developing the novel anti-HIV drugs. Constructing the QSAR model of the HIV-1 IN pyrimidones inhibitors can help for understanding of the structural factors. In this paper, the CoMFA software was used to calculate topological descriptors, polarizable descriptors and hydrophilic descriptors and other structural parameters for 68 compounds. The structural parameters were selected as inputs of support vector machine （SVM） to establish the non-linear regression model. Results show that the SVM algorithm combined with QSAR can establish the forecasting model for inhibitors, and provide biological information for the design of anti-HIV drugs.

关 键 词：支持向量机 HIV-1整合酶抑制剂 定量构效关系 嘧啶酮 

分 类 号：R914.2[医药卫生—药物化学]