Skip to main content

Table 3 Models’ performance (root-mean-square error) on lipophilicity database

From: A self-attention based message passing neural network for predicting molecular lipophilicity and aqueous solubility

Dataset (size)ModelRMSE
Lipophilicity (4200)RF0.824 ± 0.041
MPN (Deepchem)a0.630 ± 0.059
MPN (Deepchem)b0.652 ± 0.061
MPN0.630 ± 0.059
SAMPN0.579 ± 0.036
Multi-MPN0.594 ± 0.039
Multi-SAMPN0.571 ± 0.032
Water solubility (1311)RF1.096 ± 0.092
MPN (Deepchem-1128)a0.580 ± 0.030
MPN (Deepchem)b0.676 ± 0.022
MPN0.694 ± 0.050
SAMPN0.688 ± 0.057
Multi-MPN0.674 ± 0.074
Multi-SAMPN0.661 ± 0.063
  1. Italics represents the best performance in the results
  2. aValues were reported in [16]. In the lipophilicity prediction, we use the same dataset with Deepchem. In the water solubility prediction, our used dataset is larger than Deepchem used (1128 molecules)
  3. bValues were calculated from the same data and the same stratified cross-validation protocol in our work