Skip to main content

Table 1 The results of the optimization tasks

From: MERMAID: an open source automated hit-to-lead method based on deep reinforcement learning

Method Penalized LogP QED
1st 2nd 3rd Validity 1st 2nd 3rd Validity
ZINC −9.41 0.285
Single −2.20 −2.39 −2.53 29.8% 0.681 0.671 0.666 62.6%
Multi 11.33 11.20 11.10 31.8% 0.920 0.915 0.912 77.0 %
  1. The mean of the top 3 highest scored Penalized LogP and QED scores of the validation set and the validity rate are described