Table 1. Regression and Classification Task Data Sets Used in the Researcha.
| database | n | min | max | units | reference |
|---|---|---|---|---|---|
| log P | 5574 (8199) | –4.64 | 8.27 | (26) | |
| BACE-1 pIC50 | 285 (1513) | 2.699 | 10.523 | μM (log) | (27,28) |
| solubility | 803 (1128) | –11.6 | 1.58 | mols per liter (log) | (28) |
| lipophilicity | 2237 (4200) | –1.50 | 4.48 | (28) | |
| ionization energy | 1575 (2147) | 1.04 | 13.94 | eV | (29) |
| melting pointb | 7316 (10765) | –196.0 | 492.5 | °C | (10,30) |
| database | n | class ratio | reference |
|---|---|---|---|
| BBBP | 1089 (2053) | 161/928 | (31) |
| Ames mutagenicity | 3801 (6512) | 1779/2022 | (32) |
| hERG | 2452 (6795) | 1068/1385 | (33) |
| ClinTox | 707 (1478) | 56/651 | (28) |
a
The number of entries (n) is the number of molecules that lie within the applicability domain. The number in parentheses is the original number of molecules in the database.
b
Melting point database is a combination of two other databases.