Figure 7.
Examples of the pose predictions for some compounds from the Deaton et al. dataset. (a) The docked model for compound 1am using Vina (shown in blue) compared to the native pose of NAD observed in the structure 2O3U. (b) Top-scoring docked poses from twenty-fix random compounds from Vina, Glide, and rDock. Despite the similar scaffolds, Glide and rDock had difficulties finding similar poses, which may have contributed to the higher variance in the predicted affinities.
