2D-QSAR Modeling of Chalcone Analogues as Angiotensin Converting Enzyme Inhibitor

Abstract

Targeting angiotensin-converting enzyme (ACE) comes out to be an effective mechanism for controlling hypertension. Two-dimensional quantitative structural activity relationship models were generated to predict the ACE inhibitory activity of chalcone analogs. The genetic algorithm-multiple linear regression models (GA-MLR) approach was used to generate highly predictive models using straightforwardly interpretable Py, Estate, Alvadesc, and Padel descriptors. Application of Intelligent consensus modeling confirms that model-2 is statistically robust (R2 tr = 0.66, Q2 LOO = 0.5621) with good external predictivity (Concordance Correlation Coefficient, CCCex = 0.9109, Q2-F1 = 0.85818, Q2-F2 = 0.85782 and Q2-F3 = 0.88489). Novel analogs designed according to the synthetic route considering structural requirements indicated by the model were found to be satisfactory and could be considered for synthesis and subsequent screening.