Polański J
Department of Organic Chemistry, Institute of Chemistry, University of Silesia, Katowice, Poland.
Acta Biochim Pol. 2000;47(1):37-45.
A novel scheme for modeling 3D QSAR has been developed. A method involving multiple self-organizing neural network adjusted to be analyzed by the PLS (partial least squares) analysis was used to model 3D QSAR of the selected colchicinoids. The model obtained allows the identification of some structural determinants of the biological activity of compounds.
已开发出一种用于3D QSAR建模的新方案。使用一种涉及根据偏最小二乘法（PLS）分析进行调整的多个自组织神经网络的方法，对所选秋水仙碱类化合物进行3D QSAR建模。所获得的模型能够识别化合物生物活性的一些结构决定因素。
