Prediction of Lipophilicity of some Quinolone Derivatives by using Quantitative Structure-Activity Relationship

Quantitative structure activity relationship (QSAR) was used to study the partition coefficient of some quinolones and their derivatives.

These molecules are broad-spectrum antibiotic pharmaceutics. First, data were divided into two categories of train and test (validation) sets using a random selection method. Second, three approaches, including stepwise selection (STS) (forward), genetic algorithm (GA), and simulated annealing (SA) were used to select the descriptors, to examine the effect feature selection methods. To find the relation between descriptors and partition coefficient, multiple linear regression (MLR), principal component regression (PCR) and partial least squares (PLS) were used.

QSAR study showed that both regression and descriptor selection methods have a vital role in the results. Different statistical metrics showed that the MLR-SA approach with (r2=0.96, q2=0.91, pred_r2=0.95) gives the best outcome.

The proposed expression by the MLR-SA approach can be used in the better design of novel quinolones and their derivatives.