Douali L, Villemin D, Cherqaoui D
Département de Chimie, Faculté des Sciences Semlalia BP 2390 Université Cadi Ayyad, Marrakech, Morocco.
Curr Pharm Des. 2003;9(22):1817-26. doi: 10.2174/1381612033454423.
Among the non-nucleoside reverse transcriptase inhibitors, 1-[2-hydroxyethoxy-methyl]-6-(phenylthio) thymine (HEPT) derivatives have proved to be potent and selective inhibitors of human immunodeficiency virus (HIV-1). They are able to completely suppress virus replication in cell cultures. The quantitative structure-activity relationships (QSAR) try to describe the association between biological activities of a group of congeners and their molecular descriptors. In this paper, recent works on the application of neural networks (NN) and multiple regression analyses to quantitative structure-anti-HIV activity of HEPT derivatives are reviewed. NN have their origins in efforts to reproduce computer models of the information processing that takes place in the brain. They have found application in a wide variety of fields, such as image analysis of facial features, stock market predictions, etc. Application of the NN methods to problems in chemistry and biochemistry has rapidly gained popularity in recent years. We briefly describe a methodology for designing NN for QSAR and estimating their performances, and apply this approach to the prediction of anti-HIV activity of HEPT. The predictive power of the NN used is compared with that of other statistical methods.
在非核苷类逆转录酶抑制剂中，1-[2-羟乙氧基甲基]-6-(苯硫基)胸腺嘧啶（HEPT）衍生物已被证明是人类免疫缺陷病毒（HIV-1）的强效和选择性抑制剂。它们能够完全抑制细胞培养中的病毒复制。定量构效关系（QSAR）试图描述一组同系物的生物活性与其分子描述符之间的关联。本文综述了最近关于神经网络（NN）和多元回归分析在HEPT衍生物定量结构-抗HIV活性应用方面的研究工作。神经网络起源于对大脑中信息处理的计算机模型进行再现的努力。它们已在广泛的领域中得到应用，如面部特征的图像分析、股票市场预测等。近年来，神经网络方法在化学和生物化学问题中的应用迅速受到欢迎。我们简要描述了一种用于设计QSAR神经网络并评估其性能的方法，并将此方法应用于HEPT抗HIV活性的预测。将所使用的神经网络的预测能力与其他统计方法的预测能力进行比较。
