DFT based QSAR study on quinolone-triazole derivatives as antibacterial agents

QSAR modeling was performed on 39 quinolone-triazole derivatives against gram-positive Staphylococcus aureus and gram-negative Pseudomonas aeruginosa bacteria. The molecular structures were optimized using the DFT/B3LYP method and 6-31 G basis set. Molecular descriptors were extracted using quantum mechanical calculations. The hierarchical cluster analysis was performed for a rational subset division. The initial dataset was divided into calibration and validation sets, and modeling was done by stepwise MLR method for each of the two bacteria. Internal and external validation methods confirmed the robustness and predictability of the obtained models. According to the obtained model for S. aureus (R-2 = 0.889, R-ext(2) = 0.938, Q(LOO)(2) = 0.853), the four descriptors- partial atomic charges for the N1 atom in triazole and C7 of the quinolone nucleus, 4-carbonyl bond length, and C-13-NMR chemical shift of 3-carboxylic acid- were found to be the descriptors controlling the activity. According to the obtained model for P. aeruginosa (R-2 = 0.957, R-ext(2) = 0.923, Q(LOO)(2) = 0.909), the O atom's partial charge in carbonyl, LUMO-HOMO energy gap, and logP were found to be the descriptors having the highest correlation with the antibacterial activity. Finally, some new compounds with higher activities were designed and proposed.

Automated summary

QSAR modeling was performed on 39 quinolone-triazole derivatives against gram-positive Staphylococcus aureus and gram-negative Pseudomonas aeruginosa bacteria. Molecular descriptors were extracted using quantum mechanical calculations, and hierarchical cluster analysis was used for subset division. The resulting models were internally and externally validated, highlighting specific descriptors that control antibacterial activity for each bacterium. New compounds with higher activities were proposed based on the models.