Convenient nonlinear model for predicting the tissue/blood partition coefficients of seven human tissues of neutral, acidic, and basic structurally diverse compounds.

In this work, the tissue/blood partition coefficients of seven human tissues were calculated using a nonlinear regression analysis. The dataset contained 80 structurally diverse compounds distributing into the brain, kidney, muscle, lung, liver, heart, and fat, whose acidic and basic properties were also considered by introducing the three possible forms of the compound in the human body (neutral, cationic, and anionic forms). A total of 248 data points were there in the training set (eq 5: r = 0.877, s = 0.352; eq 6: r = 0.869, s = 0.362) and 49 data points in the testing set (eq 5: r = 0.844, s = 0.342; eq 6: r = 0.860, s = 0.311). It was also concluded that the same state (neutral, cation, and anion) of a compound has essentially identical partition coefficients between the same tissue composition and the blood in these tissues. Only the different content of the three tissue compositions (lipid, protein, and water) lead to the different partition coefficient in different tissues, which offered a significant conclusion for the drug's distribution research.