Predicting the free energies of complexation between cyclodextrins and guest molecules: Linear versus nonlinear models

Purpose. In the present paper, linear and nonlinear models for complexation of α- β- and γ-cyclodextrin with guest molecules are developed, with the aim of free energy prediction and interpretation of the association process. Methods. Linear and nonlinear regression is used to correlate experimental free energies of complexation with calculated molecular descriptors. Molecular modeling supports the interpretation of the results. Results. Highly predictive models are obtained, although the structural variability of the compounds used for their deduction is large, reaching from synthetic heterocycles to steroids and prostaglandins. Conclusions. The scaled regression coefficients give insight to the complexation mechanisms, which appear to be different for the three types of cyclodextrins.