The 2D-QSAR study, Drug likeness and in-silico ADMET prediction of about 3,5-diaryl-1H-pyrazole derivatives as multifunctional agents for the treatment of Alzheimer's disease.

Abstract

In this research, the activity of 20 3,5-diaryl-1H-pyrazole derivatives was used as acetylcholinesterase inhibitors for the control of Alzheimer's disease (AD), an integrated 2D-QSAR quantitative structure-activity relationship calculation technique. Density functional theory (DFT) calculation with Becke's three-parameter hybrid method and Lee-Yang Parr's B3LYP function using the 6-31G (d) basis set is applied to calculate electronic descriptors, for topological descriptors ChemSketch and MarvinSketch programs are used. The dataset was randomly divided into training sets (15 compounds) which were used to generate the QSAR model and test sets (5 compounds) which were used to evaluate the predictive ability of the QSAR model. Many statistical coefficients were thus used to select the best model. (N =15; R=0.91; R2= 0.83; F = 17.549; MSE = 0.032; Adjusted R= 0.78; p-value<0.00017) The MLR proposed QSAR model was validated internally and externally by several criteria, namely the randomization test and the Golbreich-Tropsha criteria. The applicability domain of the proposed model was applied using the Williams diagram to determine which compounds are outside this domain Based on the better proposed QSAR model, new compounds with higher acetylcholinesterase inhibition ability were theoretically designed. As such, the drug-likeness and ADME prediction performed almost showed compliance with Lipinski's rule, and the molecules were shown to be good in terms of absorption, distribution, metabolism, and excretion in general. These results may provide useful theoretical references for future experimental work.