Quantitative structure-activity relationship (QSAR) of tacrine derivatives against acetylcholinesterase (AChE) activity using variable selections.

A diverse approach to the quantitative structure-activity relationship (QSAR) of tacrine derivatives against acetylcholinesterase (AChE) activity was studied using variable selections of stepwise multiple linear regression (MLR), genetic algorithm (GA)-MLR, and simulated annealing (SA)-MLR. AChE activity (logRA) of tacrine derivatives was expressed with acceptable explanation (95.5-95.9%) and good predictive power (94.5-95.2%), respectively, in the models. The best equation was obtained from simulated annealing (SA) MLR with greater explanatory capability and better prediction, with a smaller standard error than other methods. The resulting models with the given descriptors illustrate the significant roles of hydrophobic and electrostatic interaction on increasing AChE activity, but hydrophilic and topological feature of molecules were shown to decrease AChE activity.