The electron conformational genetic algorithm (EC-GA) method had been employed by distinguishing between enantiomers for the first time as a 4D-QSAR approach to reveal the pharmacophore (Pha) and to predict the bioactivity of the dipeptidyl boron compounds. The Electron Conformational Matrices of Congruity (ECMCs) were prepared for all conformers of compounds in the data set based on the quantum chemical calculations at HF/3-21 G level in an aqueous medium. The comparison of the ECMCs within the certain tolerances by the EMRE program revealed the pharmacophore for some dipeptidyl boron derivatives. For the selection of the most influential parameters on the activity and the calculation of theoretical activities, the genetic algorithm with the non-linear least square method was used. The final model was validated by the cross-validation method with the division of the data set into training and test items. The 12-parameter model gave excellent statistical results (R2training = 0.850, R2test = 0.809, q2 = 0.755, q2ext1 = 0.776, q2ext2 = 0.759, q2ext3 = 0.735, CCCtr = 0.922, CCCtest = 0.846, CCCall = 0.905). Because of the inexistence of 4D-QSAR studies on the dipeptidyl boron derivatives and the stereoisomerism effect on the biological activity was examined for the first time for these compounds, this study plays an important role in the development of new boron-containing compounds.
Keywords: 4D-QSAR; Dipeptidyl boron; Electron conformational-genetic algorithm method; Pharmacophore; Stereoisomerism.
Copyright © 2019 Elsevier Ltd. All rights reserved.