QSAR studies of bioconcentration factors of polychlorinated biphenyls (PCBs) using DFT, PCS and CoMFA.

Abstract The bioconcentration factors (BCFs) of 58 polychlorinated biphenyls (PCBs) were modeled by quantitative structure–activity relationship (QSAR) using density functional theory (DFT), the position of Cl substitution (PCS) and comparative molecular field analysis (CoMFA) methods. All the models were robust and predictive, and especially, the best CoMFA model was significant with a correlation coefficient ( R 2 ) of 0.926, a cross-validation correlation coefficient ( Q 2 ) of 0.821 and a root mean square error estimated ( RMSE ) of 0.235. The results indicate that the electrostatic descriptors play a more significant role in BCFs of PCBs. Additionally, a test set was used to compare the predictive ability of our models to others, and results show that our CoMFA model present the lowest RMSE . Thus, the models obtain in this work can be used to predict the BCFs of remaining 152 PCBs without available experimental values.