Fig. 5

Cross-validated performance (RMSE) of the top performing model where the lattice energy (LE) descriptor was incorporated (LHS), compared to the corresponding model which didn’t include the lattice energy descriptor (RHS): dataset = Avdeef_ExDPs_CS_True. The distributions of cross-validated results are presented as a boxplot, with whiskers extending 1.5 times the interquartile range beyond the upper and lower quartiles, with the arithmetic mean superimposed as a black circle. The presence of a star denotes an apparently statistically significant difference in cross-validated mean RMSE