Prioritization scheme for quantitative structure-permeability relationship models to predict dermal absorption of chemicals

Skin permeability coefficient (k_p) is routinely used to quantify the movement of chemicals across the skin. Log octanol-water partition coefficient (log K_ow) and molecular weight (MW) are often incorporated into skin permeation models to generate the skin permeability coefficient. Given that the same dataset is used to estimate skin permeation, novel approaches are required to obtain targeted and accurate results. The main goal of this study is to identify a prioritization scheme for quantitative structure–permeability relationships (QSPRs) when using two molecular descriptors, log K_ow and MW. A second goal is to determine whether classification based on functional groups and structural similarities enhances the existing QSPR models. Ten QSPR models using log K_ow and MW were reviewed to identify the predictive ability of k_p by using a comprehensive dataset. This dataset was filtered to identify molecules with structural and functional group similarities. The resulting dataset was then subject to the QSPRs used in the preceding analysis to demonstrate improvements in their predictive abilities. By comparing the k_p predictions of the QSPRs to measured k_p values, we were able to devise a systematic approach to improve the predictive ability of the QSPRs. Using the proposed hierarchical approach, researchers can select an appropriate QSPR model to accurately predict the dermal permeability coefficient of a given chemical compound. Such predictions can be a viable alternative to experimentation, which can be resource-intensive.