Skip to main content

Advertisement

Table 6 Results of the constrained optimization for Junction Tree Variational Autoencoder [30] (JT-VAE), Graph Convolutional Policy Network [35] (GCPN) and Mol-CycleGAN

From: Mol-CycleGAN: a generative model for molecular optimization

\(\delta \)JT-VAEGCPNMol-CycleGAN
ImprovementSimilaritySuccess (%)ImprovementSimilaritySuccess (%)ImprovementSimilaritySuccess (%)
01.91 ± 2.040.28 ± 0.1597.54.20 ± 1.280.32 ± 0.12100.08.30 ± 1.980.16 ± 0.0999.75
0.21.68 ± 1.850.33 ± 0.1397.14.12 ± 1.190.34 ± 0.11100.05.79 ± 2.350.30 ± 0.1193.75
0.40.84 ± 1.450.51 ± 0.1083.62.49 ± 1.300.47 ± 0.08100.02.89 ± 2.080.52 ± 0.1058.75
0.60.21 ± 0.750.69 ± 0.0646.40.79 ± 0.630.68 ± 0.08100.01.22 ± 1.480.69 ± 0.0719.25
  1. The biggest improvements across all methods are italicized