Identification of promising inhibitory heterocyclic compounds against acetylcholinesterase using QSAR, ADMET, biological activity, and molecular docking.

Heterocyclic compounds exert diverse functions, especially acetylcholinesterase (AChE) inhibition. Thus, identifying the association between their detailed structures and biological activities is important to the development of novel medications for Alzheimer's disease (AD) treatment. In this study, diverse sets of 120 potent and selective heterocyclic compounds (-log[the half‑maximal inhibitory concentration] (pIC50) values ranged from 8.01 to 12.50) were used to develop quantitative structure-activity relationship (QSAR) models using multiple linear regression (MLR), multiple nonlinear regression (MNLR), Bayesian model average (BMA), and artificial neural network (ANN) models. The models' robustness and stability have been assessed using both internal and external methodology. ANN outperforms MLR, MNLR, and BMA according to external validation. The molecular descriptors incorporated into the model were in satisfactory correlation with the AChE receptor-ligand complex X-ray structures, making the model interpretable and predictive. Three selected compounds exert drug-like characteristics (pIC50 values ranged from 11.01 to 11.17). The binding affinity between the optimal compounds and the AChE receptor (RCSB ID 3LII) ranged from - 7.4 to - 8.8 kcal/mol. Remarkably, the pharmacokinetics, physicochemical properties, and biological activities of compound 25 (C23H32N2O2, PubChem CID 118727071, pIC50 value = 11.17) were found to be consistent with its therapeutic effects in AD due to its cholinergic and non-toxic nature, non-P-glycoprotein, high gastrointestinal absorption, and capability to penetrate the blood-brain barrier.