Table 1. A summary of the computational and experimental results.
V1 | V2 | V3 | V4 | |
IC50 | 2.16±1.20 | 1.53±1.17 | 8.36±1.71 | 1.59±1.10 |
EC50 | >100 | >100 | >100 | 2.16±0.25 |
AutoDockCrystal | −11.8 | −11.3 | −10.7 | −11.8 |
RankCrystal | 11 | 20 | 31 | 25 |
AutoDockEnsemble | −11.9±1.4 | −11.9±1.4 | −10.2±1.0 | −12.8±1.6 |
RankEnsemble | 4 | 3 | 12 | 1 |
% Expected Pose | 33% | 33% | 24% | 18% |
AutoDockEnsemble/Expected | −13.3±1.2 | −12.6±1.1 | −11.6±0.5 | −14.9±0.6 |
AutoDockEnsemble/Unexpected | −11.2±0.9 | −11.6±1.4 | −9.8±0.6 | −12.3±1.4 |
AutoDockEnsemble/Best | −15.3 | −13.9 | −12.1 | −15.6 |
IC50 and EC50 are measures (in µM) of the inhibition of REL1 activity and parasite growth, respectively; AutoDockCrystal is the predicted binding energy, in kcal/mol, to the crystal structure; RankCrystal is the rank of the ligand when all 588 compounds are ordered by their respective AutoDockCrystal values; AutoDockEnsemble is the average predicted binding energy to the 33 representative protein-receptor conformations obtained via QR factorization, plus or minus the standard deviation; RankEnsemble is the rank of the ligand when the top 45 compounds are ordered by their respective AutoDockEnsemble values; % Expected Pose is the percentage of the 33 representative protein structures amenable to deep-pocket binding, in which the naphthalene core is docked deep into the binding pocket; AutoDockEnsemble/Expected is the average predicted binding energy when only those members of the ensemble amenable to deep-pocket binding are considered; AutoDockEnsemble/Unexpected is the average predicted binding energy when only the remaining members of the ensemble are considered; and AutoDockEnsemble/Best is the predicted binding energy of the ligand to the “optimal protein conformation” from the ensemble. All predicted energies are in kcal/mol.