Application of neural networks using the volume learning algorithm for quantitative study of the three-dimensional structure-activity relationships of chemical compounds

A volume learning algorithm for artificial neural networks was developed to quantitatively describe three-dimensional structure-activity relationships using as an example N-benzylpiperidine derivatives. The new algorithm comb ines two types of neural networks, the Kohonen and the feed-forward artific ial neural networks, which are used to analyze the input grid data generate d by the comparative molecular field approach. Selection of the most inform ative parameters using the algorithm helped reveal the most important spati al properties of the molecules, which affect their biological activities. C luster regions determined using the new algorithm adequately predicted the activity of molecules from a control data set.