Fig. 5

Validating the model for the known in-actives for the receptor 1BNU. The figure presents the comparison of the docking enrichment in green and the predicted enrichment in magenta for the known in-active compounds. The comparison was used to validate the performance of the 1BNU receptor model for the known in-active compounds. The docking enrichment was created by comparing the docking scores of the known in-actives and the docking scores of the external validation set. Similarly the predicted enrichment was created by comparing the predicted p-values for the known in-actives and the p-values for the external validation set. AUC was also calculated and reported in the figure for comparison. Overall, the 1BNU model performed well and the predicted enrichment was comparable to the docking enrichment. The green line for the docking enrichment, which was below the random grey line, also confirms the validity of the virtual screening evaluation