A conceptual DFT and information-theoretic approach towards QSPR modeling in polychlorobiphenyls

Quantitative structure-property relationship (QSPR) of various chlorine substituted biphenyl systems on the basis of linear and multi-linear regression (MLR) analysis is presented in this study. The determination of lipophilicity (log K-OW) of the selected 133 polychlorobiphenyl (PCB) congeners is carried out taking the experimental log K-OW as the dependent variable and the conceptual density functional theory (CDFT) and information theory (IT) based descriptors (global electrophilicity index (omega), its square term (omega(2)), and Shannon entropy (S-S), GBP entropy (S-GBP)) as independent variables. These are used to map the relationship between experimental log K-OW and predicted log K-OW. The best model is obtained using CDFT descriptor (omega) along with IT quantities (S-S, S-GBP) when combined linearly. The results show a very good coefficient of determination value (R-2 = 0.9261) along with a high internal predicting ability (R-CV(2) = 0.9208) which indicates the importance of the mentioned descriptors for the quantitative structure-property analysis of selected PCBs.

Automated summary

This study presents the quantitative structure-property relationship (QSPR) of various chlorine substituted biphenyl systems based on linear and multi-linear regression (MLR) analysis. The lipophilicity (log K-OW) of 133 selected polychlorobiphenyl (PCB) congeners was determined using experimental log K-OW as the dependent variable and conceptual density functional theory (CDFT) and information theory (IT) based descriptors as independent variables. The relationship between experimental and predicted log K-OW was mapped using these descriptors. The best model was obtained by combining CDFT descriptor (omega) with IT quantities (S-S, S-GBP) linearly, showing a high coefficient of determination value (R-2 = 0.9261) and excellent internal predicting ability (R-CV(2) = 0.9208), indicating the importance of the mentioned descriptors for quantitative structure-property analysis of selected PCBs.