J. Comput.-Aided Mol. Des.

The observed 5-HT1A and alpha(1)-adrenergic receptor (alpha(1)-AR) receptor binding properties of a series of 23 thienopyrimidinones were used to develop HASL 3D-QSAR models. A single, low energy conformer of the most active analogue in the series, which was consistent with NMR structural studies, was chosen as a template molecule. Alignments of all the molecules to the template were provided by an Amber/MM2 superposition force field. In this manner, each molecule was represented by five separate low energy conformers which were subsequently used in the generation of HASL 3D-QSAR models. Models derived from multiple conformers were found to exhibit enhanced predictivity compared to models based on single, low energy conformers. In addition, the use of contour imaging of HASL multi-conformer model interactions was found to lead to a more consistent interpretation of those molecular features most significant for 5-HT1A receptor binding.The observed 5-HT1A and alpha(1)-adrenergic receptor (alpha(1)-AR) receptor binding properties of a series of 23 thienopyrimidinones were used to develop HASL 3D-QSAR models. A single, low energy conformer of the most active analogue in the series, which was consistent with NMR structural studies, was chosen as a template molecule. Alignments of all the molecules to the template were provided by an Amber/MM2 superposition force field. In this manner, each molecule was represented by five separate low energy conformers which were subsequently used in the generation of HASL 3D-QSAR models. Models derived from multiple conformers were found to exhibit enhanced predictivity compared to models based on single, low energy conformers. In addition, the use of contour imaging of HASL multi-conformer model interactions was found to lead to a more consistent interpretation of those molecular features most significant for 5-HT1A receptor binding