QSPR of n‐Octanol/Water Partition Coefficient of Nonionic Organic Compounds Using Extended Topochemical Atom (ETA) Indices

Abstract

AbstractConsidering n‐octanol/water partition coefficient as an important chemical property in modeling the fate and persistence of chemicals, recently introduced Extended Topochemical Atom (ETA) indices have been used to model n‐octanol/water partition coefficient data of 122 nonionic organic compounds which are highly persistent in the environment and reported to bioconcentrate considerably in lipid tissues. In deriving the models, principal component Factor Analysis (FA) followed by Multiple Linear Regression (MLR), stepwise regression, Partial Least Squares (PLS), and Principal Component Regression Analyses (PCRA) were applied as the statistical tools. The model development process was repeated with non‐ETA (topological and physicochemical) descriptors and a combination set comprising both the ETA and non‐ETA descriptors. The models with ETA indices suggested negative contributions of groups capable of hydrogen bonding and/or polar interactions and positive contributions of volume and degree of halogen substitution to the partition coefficient. Finally, we discuss validation of QSPR models by dividing the dataset into training and test sets based on different strategies, e.g., random division, sorted log Kow data and K‐means clusters for the factor scores of the original variable (ETA) matrix without the response property values. The ETA (FA‐MLR and stepwise regression) models were also applied on a purely external dataset (n=35) and acceptable predictive r2 values (0.582 and 0.651 respectively) were obtained. The results suggest that ETA parameters are sufficiently rich in chemical information to encode the structural features contributing to the n‐octanol/water partition coefficient of nonionic organic compounds and thus these merit further assessment to explore their potential in QSAR/QSPR/QSTR modeling.