A 3D-QSAR of N-Substituted 4-Amino3-3-Dialkyl-2 (3H)-Furanone GABAA Receptor Modulators Based on Receptor Surface Analysis

A hypothetical receptor surface model has been generated for an allosteric receptor site, "lactone site", on the GABAA receptor using a set of N-substituted 4-amino-3, 3-dialkyl-2 (3H)-furanone GABAA receptor modulators (I). The activity data of each of the molecules was used as a weight in building of receptor surface. The study produced reasonably good predictive model with good conventional and cross-validated r2 with leave-two-out method (0.998 and 0.943 respectively) in the training set and also in external data set validation (r2pred =0.960). Most of the points selected for QSAR to conduct RSA are in the vicinity of the groups in position 3 and 4 and not in the vicinity of carbonyl group and oxygen atom region, predicted as favourable site for binding of molecules. These findings and the lack of direct correlation between interaction energy and biological activity led to infer that the compounds may function following "Message-Address" concept.