Predicting blood–brain barrier penetration from molecular weight and number of polar atoms

A simple three-descriptor model to predict blood–brain barrier is derived from a training set of 78 compounds: log BB = −9.880 × 10−6MW2 + 7.339 × 10−3 MW–0.2268npol−0.1143 (n = 78, r2 = 0.74), where log BB is the logarithm of the ratio of the steady-state concentration of the compound in the brain to concentration in the blood, MW is the molecular weight, npol is the number of polar atoms (oxygen, nitrogen, and attached hydrogen), n is the number of compounds, and r is the correlation coefficient. The model is validated through use of leave-one-out procedure and an external test set (25 compounds). The model is suitable for the rapid prediction of the blood–brain barrier penetration of drug candidates because of its predictive ability and simplicity.