辨识药物定量构效关系的模糊神经网络方法研究

提出一种基于遗传算法的新型模糊神经网络方法,用于计算Benzodiazepines(BZs)类药物的定量构效关系.这类模糊神经网络综合了神经网络、遗传算法与模糊逻辑的各自优势,具有优良的定量构效关系辨识能力,其学习速度较快,不易陷入局部最小区域;网络知识以模糊语言变量的形式加以表达,不仅易于理解,而且能有效地利用已有的专家经验.一旦通过学习获得规律后,不仅能很好地预测化合物的活性,还能对后续的药物分子设计提供有益的理论指导.

Studies on Quantitative Structure-activity Relationships of Benzodiazepines Using Fuzzy Neural Networks

In this paper, a new fuzzy neural network based on genetic algorithms is proposed for quantitative structure-activity relationship (QSAR) studies of benzodiazepines. The method based on GA+FL +-NN allows supervised learning of fuzzy rules from significant examples and is affected unsusceptibly by the problem of local extremes. The network′ s knowledge base has a linguistic representation. This makes it easy for pharmaceutical chemists to understand and interpret. It is possible to introduce current knowledge acquired by researchers simply by adding one or more fuzzy rules to the network′ s knowledge base. Once the fuzzy knowledge base extracted from examples, it can predict the pharmacological activity of compounds at a high precision. The obtained fuzzy rules can also provide useful guidelines for synthesizing new compounds with a high pharmacological activity.