From: Transformer-based molecular optimization beyond matched molecular pairs
Datasets | Training (2000-2017) | Validation (2018) | Test (2019-2020) |
---|---|---|---|
MMPs | 2,287,588 | 143,978 | 166,582 |
Similarity (\(\ge\)0.5) | 6,543,684 | 418,180 | 475,070 |
Similarity ([0.5,0.7)) | 4,543,472 | 286,682 | 327,606 |
Similarity (\(\ge\)0.7) | 2,000,212 | 131,498 | 147,464 |
Scaffold | 2,850,180 | 171,914 | 199,786 |
Scaffold generic | 4,127,058 | 255,580 | 289,034 |