Benchmarking docking and scoring protocol for the identification of potential acetylcholinesterase inhibitors

Acetylcholinesterase (AChE) plays a crucial role in nerve impulse transmission at cholinergic synapses by rapid hydrolysis of the neurotransmitter acetylcholine (ACh). AChE has become an important drug target because partial inhibition of AChE results in modest increase in ACh levels that can have therapeutic benefits, thus AChE inhibitors have proved useful in the symptomatic treatment of Alzheimer's disease. To establish an effective docking protocol for virtual screening of AChE, a comparative molecular docking study was performed. For this purpose six docking/scoring approaches (AutoDock, FlexX, MOE, Surflex-Dock, GOLD and FRED) were compared to determine their ability to reproduce the binding poses in twenty six complexes of AChE. Docking accuracy was evaluated by calculating the RMSD of the docked complexes. FRED was found to be the best in reproducing the experimental pose by placing it near the top of its ranking. The performance of scoring functions was evaluated by identifying known actives out of large database of inactive compounds. A dataset of 5000 "drug like" decoys were retrieved from NCI database and docked into the binding site of AChE with six known inhibitors using FRED in combination with five scoring functions, i.e., Chemgauss2, Chemgauss3, ChemScore, Shapegauss and PLP. The poses obtained by FRED were re-scored using GOLD score, ChemScore and ASP as implemented in GOLD while G_Score, D_Score, ChemScore and PMF as implemented in the CScore module of SYBYL7.3. D_Score presented significantly better enrichment than others and 50% of the active inhibitors were identified in top 20% of the ranked database.