Table 1 Structural transformations—dataset sizes
From: Mol-CycleGAN: a generative model for molecular optimization
Dataset | Halogen moieties | Aromatic rings | Bioisosteres | \(\text {CF}_3\) |
|---|---|---|---|---|
\(X_{\text {train}}\) | 75,000 | 80,000 | 12,454 | 20,000 |
\(X_{\text {test}}\) | 86,899 | 18,220 | 800 | 800 |
\(Y_{\text {train}}\) | 75,000 | 80,000 | 8321 | 8799 |
\(Y_{\text {test}}\) | 12,556 | 43,193 | 800 | – |