LAUSR

Four‐dimensional quantitative structure‐activity relationship (4D‐QSAR) models were developed to predict biological activity of 1,4‐naphthoquinones derivatives tested against human HL‐60 leukemic cells, in order to better investigate the action mode of these compounds. Quinones can generate reactive oxygen species (ROS) through the activation by the cytochrome P450 and P450 reductase enzymes acting as anticancer agents. Molecular dynamics (MD) simulations were performed in the 3D optimized geometries, and the field descriptors were calculated. Partial least squares (PLS) regression method was applied to build the QSAR model, which presented the following statistics, with two factors and explaining 51.11% of total variance: R2 = 0.83; SEC = 0.28; Q2 = 0.77; SEV = 0.31. For external validation, the results were R2pred = 0.76 and SEP = 0.30. Among the nine Coulomb (C) and Lennard‐Jones (LJ) descriptors selected by the model, one of them, C13838, is located close to quinone oxygens involved in the production of radical anions (O2−·) and to hydroxyl in position 5 that may stabilize catalytically important water molecules. The negative LJ descriptors around R1 and R2 substituents might indicate that apolar substituents in these regions are unfavorable to the activity. Coulomb descriptors located at the vicinities of the substituent R2 gave information about the bioactive conformation.