ABSTRACT
In this research, a combined method of ligand-based pharmacophore (LBP), structure-based pharmacophore (SBP), and molecular docking was applied for virtual screening potential ATP-sensitive potassium channel (KATP) openers from Chinese herbs. LBP models were generated by 3D-QSAR pharmacophore(hypogen) program, based on the training set composed of 48 KATP agonists. The best LBP model consisted of one hydrogen-bond acceptor, one hydrogen-bond donor, one hydrophobic feature, one aromatic ring and five excluded volumes. Besides, the correlation coefficient of training set and test set, N, and CAI value of the model were 0.876 4, 0.705 8, 3.304, and 2.616 respectively. Meanwhile, SBP models were also generated based on a 3D structure of KATP (PMID: PM0079770). The best SBP model consisted of six hydrogen-bond acceptors, eight hydrogen-bond donors, seven hydrophobic features and eighteen excluded volumes. The corresponding N and CAI value were 2.200 and 2.017. Then, the best LBP model and SBP model were applied to identify potential KATP openers from Traditional Chinese Medicine Database(TCMD), respectively. 349 hits were obtained after analyzed by drug-likeness rules. Moreover, 12 compounds with high docking scores were reserved after molecular docking evaluation. Interestingly, part of the results had been verified as hypotensive active ingredients by literatures. Therefore, this study uncovers a specific target effect contained in TCMD, and provides candidates for new KATP openers' research.