A QSAR Study for the Prediction of Inhibitory Activity of Coumarin Derivatives for the Treatment of Alzheimer's Disease

The inhibition of acetylcholinesterase (AChE) enzyme has been used as a successful therapeutic strategy for the symptomatic treatment of Alzheimer's disease and its progression. It is also known that Coumarins, a group of naturally occurring substances in many plants, exhibit a wide range of biological activities such as AChE inhibition. In this study, we present a quantitative structure-activity relationship (QSAR) analysis to predict the inhibitory activity (IC50) of Coumarins derivatives using several statistical regression and machine learning models based on various molecular descriptors of 94 different compounds extracted by the popular Dragon software. The models include multiple linear regression (MLR), partial least squares (PLS), random forests, artificial neural networks, and support vector machine (SVM). Also, a genetic algorithm (GA) was used in combination with MLR, PLS, SVM, and ANN to find a smaller subset of the utilized descriptors. The results indicated that the GA-ANN model achieves the best IC50 prediction accuracy.

Automated summary

The study used various statistical regression and machine learning models to analyze the inhibitory activity of Coumarins derivatives, with the GA-ANN model achieving the best IC50 prediction accuracy.