Prediction of antibacterial and antifungal activity of 4 benzylideneamino benzene sulfonamides series by 2D QSAR studies

Abstract

A series of thirty one molecules substituted 4-benzylideneamino benzenesulphonamides derivatives compounds displaying variable inhibition of microbial activity were selected to develop models for estab- lishing 2D QSAR by partial least square analysis. The compounds in the selected series were char- acterized by spatial, molecular and electrotopological descriptors using QSAR model of molecular design suite (V-Life MDS 3.5). Correlations between inhibitory activities and calculated predictor variables were established through partial least square regression method. The whole dataset was divided into training set (22 compounds) and test set (09 compounds). The statistically signi cant best 2D QSAR model having correlation coefcient r2 = 0.8482and cross validated squared cor- relation coef cient q2 = 0.8094with external predictive ability of pred_r2 = 0.0947 coef cient of correlation of predicted data set (pred_r2se) 0.0879 was developed by stepwise PLSR method with the descriptors like T_2_N_5, T_2_N_3, SsssNE-index, SssOH Eindex Count, SsBrcount and Xlogp. These results should serve as a guideline in designing more potent and selective amtimicrobial molecules.