Discovering Promising Antileukemia Agents: Computational 2D-QSAR and ADMET Study on New Quinolinol Derivatives.

Drug discovery requires the continuous development of new, more active agents with fewer adverse effects. Within, In Silico approaches play a very important role in medicinal chemistry as an efficient method for the prediction of bioactive compounds. The approach used in this study is 2D-QSAR by processing data on the structural and physicochemical properties of a series of 22 quinolinol derivatives using statistical analysis methods such as Partial Least Squares (PLS), for virtual screening and choosing the most significant descriptors to have a model allowing to link the activity with the descriptors. The study using the Molecular Operating Environment (MOE) software shows that there is a significant correlation (R2 = 0.75) between biological activity and three descriptors (BCUT_PEOE_2, logS, and SlogP_VSA0) and a Root Mean Square Error (RMSE) of 0.39, with an internal validation of Q2 = 0.66 and an external validation of Rpred2 = 0.73. Also, the ADMET analysis showed good absorption, distribution, metabolism, excretion, and toxicity properties for some molecules with high biological activities. The selected 2D-QSAR model leads to the prediction In Silico of the design of new derivatives with improved activity.