QSAR Modeling and Drug-Likeness Screening for Antioxidant Activity of Benzofuran Derivatives

In order to explore the relationship between the antioxidant activity and structure of fifteen benzofuran derivatives, we carried out a QSAR study using multiple linear regression (MLR) and artificial neural network (ANN) methods. Six descriptors were used as input data (molar weight, surface area, octanol-water partition coefficient, hydration energy, highest occupied molecular orbital energy and lowest unoccupied molecular orbital energy). The electronic properties were derived at the B3LYP/6-31 G (d,p) level. Benchmarks on furan and benzofuran subunits and their comparison to the experiment showed that this level of theory is good enough. The output data correspond to the antioxidant activity as given by IC50. The predicted properties are in agreement with experimental values. Our study shows that 80% of studied molecules are in accordance with the Lipinski and Veber rules and reach the optimal lipophilicity indices. In addition, statistical analysis reveals that ANN technique with (6-2-1) architecture is more significant than MLR model. (C) 2019 Elsevier B.V. All rights reserved.