Skip to main content

Table 1 Overlap of actives (mean and standard deviation) in the top-1% search outputs from SUM-OKA and TAN searches: percentage of active molecules retrieved by both SUM-OKA and TAN ("Overlap") or only by one of these two search methods.

From: Evaluation of a Bayesian inference network for ligand-based virtual screening

  Percentage of actives
Dataset Overlap SUM-OKA TAN
MDDR 14.81 12.67 6.58 4.28 4.40 2.23
MDDR-HOM 69.21 18.48 13.39 8.84 3.58 3.75
MDDR-HET 7.05 6.36 2.38 1.86 4.71 3.75
WOMBAT 22.37 16.01 8.43 5.38 3.84 2.01