Table 3 Performance of different logP methods against the Avdeef dataset
From: JPlogP: an improved logP predictor trained using predicted data
Predictor | Performance | %Binned absolute errors | ||||
|---|---|---|---|---|---|---|
RMSE | < 0.5 | 0.5–1 | 1–1.5 | 1.5–2 | ≥ 2 | |
JPlogP-library | 0.63 | 68.91 | 20.6 | 7.12 | 2.25 | 1.12 |
LogP4Average | 0.65 | 71.91 | 16.11 | 7.49 | 3.37 | 1.12 |
AlogP (Vega) | 0.65 | 70.04 | 18.35 | 6.74 | 3 | 1.87 |
Biobyte CLogP | 0.76 | 70.41 | 17.23 | 4.12 | 3.75 | 4.49 |
XlogP3—AA | 0.77 | 69.29 | 15.36 | 8.24 | 3.75 | 3.37 |
SlogP | 0.79 | 49.06 | 34.46 | 10.49 | 4.12 | 1.87 |
Molinspiration | 0.80 | 63.30 | 20.23 | 10.49 | 3.37 | 2.62 |
JPlogP-coeff | 0.81 | 47.94 | 32.58 | 13.48 | 4.49 | 1.5 |
ACD | 0.83 | 68.17 | 19.10 | 8.24 | 1.87 | 2.62 |
KowWIN | 0.84 | 73.78 | 14.97 | 5.99 | 2.25 | 3.00 |
MlogP (Vega) | 0.85 | 67.04 | 16.85 | 6.74 | 5.24 | 4.12 |
AlogPS logP | 0.86 | 66.29 | 23.60 | 7.12 | 2.25 | 0.75 |
Myelan (Vega) | 0.89 | 65.54 | 15.73 | 9.74 | 4.49 | 4.49 |
XLogP2 | 1.05 | 56.93 | 20.22 | 8.99 | 7.12 | 6.74 |
Mannhold LogP | 1.43 | 26.22 | 24.72 | 20.97 | 13.86 | 14.23 |
AAM | 1.62 | 21.35 | 23.97 | 18.73 | 12.73 | 23.22 |
AlogP (CDK) | 2.57 | 7.87 | 10.49 | 19.1 | 14.61 | 47.94 |