Fig. 1. Representations of a brassinosteroid, showing its interaction energy with a water probe. (a) Repulsive positive energy contoured at +0.2 kcal/mol. (b) Attractive negative regions at −3.0 kcal/mol.

Fig. 2. Gibbs energy of hydration of 25 brassinosteroids. (a) Comparison between the ΔG of hydration calculated from VolSurf descriptors and by the AM1-SM2 methods (r²=0.98). (b) Plot of PLS coefficients, showing that the size of all the hydrophilic regions, the capacity factors and the polarisability are correlated with the energy of hydration. The larger the hydrophilic regions, the more negative the energy of hydration. Hydrophobic regions show the opposite pattern.

Fig. 3. Skin permeation data for 46 chemically very diverse compounds [33]. (a) Experimental vs. calculated values. (b) Experimental vs. predicted values.

Fig. 4. Coefficient plot for the skin permeation model. See text for interpretation.

Fig. 5. GRID 3D molecular fields of cortisone (upper part) and butanone (lower part) calculated with a water probe. The zones shown are the hydrophilic regions contoured at −3 kcal/mol. The two arrows represent the vectors of the integy moments calculated at this energy level.

Fig. 6. Plot of experimental vs. calculated affinity constants for the binding of drugs to the A variant of AAG (data in Table 1).

Fig. 7. Coefficient plot in the third PLS dimension for the binding of drugs to the A variant of AAG.

Fig. 8. A study of permeation acrosss Caco-2 cells [36]. (a) PLS partial weight plot to compare the VolSurf descriptors calculated here, and the PWASA descriptors of the original work [36]. The arrows indicate PWASA and VolSurf descriptor W7. (b) Experimental vs. calculated P_app obtained with a two-component VolSurf model.

Table 1. Binding association constants (−log K′) to the A variant of AAG [35]

Table 2. Boltzmann-averaged polar water-accessible surface area (PWASA) and Caco-2 cells permeability (P_app) [36], and VolSurf hydrophilic volume calculated here