2D 3D-QSAR studies on -di keto Acids Derivatives of Pyridinones as anti-HIV drugs

Quantitative structure studies 2D 3D, linear and nonlinear activity have been developed to estimate and predict the biological activity of the combinations of -di keto Acids Derivatives of Pyridinones as anti-HIV drug inhibition inhibitors. The model of the relationship between descriptors and pEC50 values of the selected molecules is described in two steps. In the first step, the 2D-QSAR method was used by selecting the descriptors using genetic algorithm and multiple linear regression, step by step, and linear models related to it. By examining the mean effects of the descriptors, it was realized that MATS7m and HATS6p have the most effect on the bioavailability of the studied compounds. Using the descriptors element, we constructed SW-MLR and GA-MLR linear models and ANN and ANFIS nonlinear models. In the second step, using the SYBYL-X software, 3D-QSAR studies were conducted to construct three-dimensional models; in the CoMFA model, the effects of spatial and electrostatic fields are taken into account, and the results are : (Q2 = 0.748,R2 = 0.906); and in relation to CoMSIA model, in which, in addition to the previous effects, hydrophilic effects, and hydrogen bond effects are taken into account and the best results are (Q2 = 0.752,R2 = 0.923). The results indicated that the CoMSIA method is more acceptable than the CoMFA method. Finally, by studying the results of the two CoMFA and CoMSIA models, the structure of some -diketo Acids Derivatives of Pyridinones compounds has been suggested for further