Three-dimensional quantitative structure-activity relationship (3D-QSAR) of 3-aryloxazolidin-2-one antibacterials

Three-dimensional quantitative structure-activity relationship (3D-QSAR) studies for 3-aryloxazolidin-2-one antibacterials were performed using the genetic function approximation algorithm. This study was performed using 60 compounds, in which the QSAR models were developed using a training set of 50 compounds, The in vitro minimum inhibitory concentration (MIC) against Staphylococcus aureus SFCO-1a was used for the study. The predictive ability of the QSAR model was evaluated by using a test set of 10 compounds. The statistical quality of the QSAR models was assessed using statistical parameters r(2), r(ev)(2) (cross-validated r(2)), r(pred)(2) (predictive r(2)) and lack of fit measure (LOF). The results obtained indicate that the antibacterial activity of the 3-aryloxazolidin-2-ones is strongly dependent on electronic factor as expressed by lowest unoccupied molecular orbital energy (LUMO), spatial factor as expressed by density and thermodynamic factors accounted for by molar refractivity and heat of Formation. The model is presently being used to design and predict new potent molecules prior to synthesis. (C) 2001 Elsevier Science Ltd. All rights reserved.