ABSTRACT
The capabilities of the probabilistic classication scheme for identication of the common reactivity pattern (COREPA) of biologically similar chemicals to model binding afnity to human estrogen receptor (hER) will be reviewed. The COREPA reactivity patterns based on the distance between nucleophilic sites allow the identication of distinct ER binding interaction mechanisms. Three-dimensional (3D) structural and parametric boundaries of different binding mechanisms are dened and used for building a categorical quantitative structure-activity relationship (QSAR) model. In addition to the traditional
10.1 Introduction .................................................................................................260 10.2 Methods ....................................................................................................... 263
10.2.1 Training Set Chemicals .................................................................. 263 10.2.2 Ligand-Binding Pocket and Interaction Mechanisms .................... 263 10.2.3 Molecular Modeling Approches .....................................................265
