From: Molecular generative model based on conditional variational autoencoder for de novo molecular design
 | Condition | Random | ||||
---|---|---|---|---|---|---|
 | Attempts | Number of valid molecules | Success rate (100/attempts, %) | Attempts | Number of valid molecules | Success rate (100/attempts, %) |
Aspirin | 28,840 | 32,567 | 0.34 | 758,199 | 711,660 | 0.014 |
Tamiflu | 15,960 | 34,696 | 0.62 | 798,183 | 741,960 | 0.013 |
Lenalidomide | 50,200 | 89,230 | 0.19 | 865,695 | 822,060 | 0.012 |
Rivaroxaban | 92,620 | 47,574 | 0.11 | 866,205 | 817,800 | 0.012 |
Pregabalin | 77,680 | 84,371 | 0.13 | 782,010 | 723,360 | 0.014 |