COMPARISON OF THREE 3D-QSAR METHODS USING A NOVEL CLASS OF MURF INHIBITORS

Abstract

MurF was considered as an attractive target for new antibacterial discovery. In this paper, three QSAR methods were employed, viz. comparative molecular field analysis (CoMFA), comparative molecular similarity indices analysis (CoMSIA) and hologram QSAR (HQSAR), to derive highly predictive QSAR models for designing novel MurF inhibitors and comparing different 3D-QSAR/alignment methods. QSAR models with high predictive ability for MurF inhibitors were successfully constructed in terms of cross-validation q2, standard error and predictive coefficient r2, which were around 0.70, 0.55 and 0.99, respectively. All the models from different methods were in good agreement with each other. Compounds with indeterminate activities were used as a test set; results showed that CoMSIA had the best predictive ability, followed by HQSAR and CoMFA. Based on these models, some key features for designing new MurF inhibitors were identified. A virtual database screen process was proposed based on the combination of these models.