An effort to discover the preferred conformation of the potent AMG3 cannabinoid analog when reaching the active sites of the cannabinoid receptors.

Most of current 3D-QSAR algorithms use alignments of compounds at the training set based on reference active ligands in the first step of the construction of the pharamacophore modeling. This first step mostly defines the success of constructed pharmacophore models. In this step, it is essential to find the bioactive conformation for solid and reliable 3D-QSAR models. Therefore, we have proceeded through different approaches for revealing the preferred conformations of Δ(8)-THC derivative AMG-3 at CB1 and CB2 receptors. In the first approach, we have applied conformational search methods in gas and in solvent phases for the ligand. The derived low energy conformers using these methodologies have been modeled through 3D-QSAR studies (first generation model). In the second approach, the low energy conformers derived from molecular docking studies have been used as input for 3D-QSAR studies (second generation model). In the current study, a new approach using MD calculations in a simulated biological environment, thus the CB receptors surrounded by a lipid bilayer environment has been used (third generation). The obtained results for different environments were compared and the approach deriving the highest statistic results was used for the generation of the novel AMG3 analogs for optimal and selective binding affinities at CB1 and CB2 receptors by the de novo drug design modeling.