Human nAChR u7 is the potential target for schizophrenia cognitive disorders, and it is meaningful to develop selective human nAChR α7 agonists for the clinical treatment of the disease. Because the crystal structure ofα7 receptor has not been resolved, ligand-based drug design strategy was took in this work. A 3D QSAR pharmacophore model was built by HypoGen method, and its quality was evaluated by cost function. Furthermore, the pharmacophore model was validated with activity prediction of test set and was cross-validated based on Fisher＇s Randomization Method. By Enrichment Factor and AU-ROC analysis, the final pharmacophore, which is consisted of one HBA, two Hydrophobic and one PosIonizable, was selected and it fitted well with the docking result of α7 homology model and the ligand. The pharmacophore is expected for the following virtual screening and lead optimization of human nAChR α7 agonists, which is important for the development and discovery of novel antipsychotics.
