Table 4.
Test set to validate the atom-based QSAR model for anti-preS1 activity. Listed are experimentally determined binding affinities (−log (IC50 M)) of ligands and predicted binding affinity by the resulting QSAR model. Error describes the difference between the experimentally determined and the predicted binding affinity. Previous descriptions of the compounds as NTCP inhibitors are referenced. As an exception, the arylmethylamino “steroid” compounds are introduced here as novel NTCP inhibitors for the first time.
| Ligand Name | Binding Affinity | Prediction | Error |
|---|---|---|---|
| Ciglitazone [35] | 6.000 | 4.570 | −1.430 |
| Betulinic acid [13] | 5.699 | 4.677 | −1.022 |
| 20,29-Dihydrobetulin [13] | 5.456 | 4.815 | −0.641 |
| 3-O-(3,3-Dimethylglutaroyl)betulinic acid [13] | 5.398 | 5.073 | −0.325 |
| 3,28-Di-O-acetyl-29-hydroxybetulin [13] | 5.097 | 4.433 | −0.664 |
| Saquinavir [36] | 5.097 | 4.349 | −0.748 |
| Emodin [39] | 5.046 | 4.845 | −0.201 |
| Ginkgolic acid 13:0 [39] | 4.959 | 4.496 | −0.462 |
| S985852 [14] | 4.854 | 4.774 | −0.080 |
| Pioglitazone [35] | 4.770 | 4.888 | 0.119 |
| Nitrendipine [37] | 4.745 | 4.390 | −0.354 |
| Glyburide [37] | 4.678 | 4.161 | −0.517 |
| Lupenone [13] | 4.658 | 4.398 | −0.259 |
| Betulonoyl dimethyl-L-aspartate [13] | 4.638 | 4.300 | −0.338 |
| 3-Oxoallobetulin [13] | 4.509 | 4.228 | −0.281 |
| Steroid 6c [31] | 4.432 | 4.040 | −0.392 |
| Steroid 7s [31] | 4.420 | 4.410 | −0.010 |
| Raloxifene [37] | 4.319 | 4.376 | 0.057 |
| Lupeol [37] | 4.292 | 4.497 | 0.204 |
| Steroid 1s [31] | 4.237 | 3.881 | −0.356 |
| 3,28-Di-O-(dihydrocinnamoyl)betulin [13] | 4.125 | 4.902 | 0.777 |
| A000289041 [14] | 4.051 | 4.708 | 0.657 |
| Nifedipine [37] | 4.018 | 4.463 | 0.445 |
| Steroid 2c [31] | 3.900 | 4.002 | 0.102 |
| Steroid 2s [31] | 3.666 | 4.075 | 0.410 |
| Steroid 2o [31] | 3.600 | 4.176 | 0.576 |