Skip to main content
. Author manuscript; available in PMC: 2016 Feb 23.
Published in final edited form as: J Chem Inf Model. 2015 Feb 6;55(2):407–420. doi: 10.1021/ci500691p

Table 4.

Comparison of enrichments for different docking and pharmacophore modeling methods.

Targets Methods EF1 EF10 EF20 AUC
HIVPR 4F SILCS-Pharma 6.7 4.1 3.6 0.78
5F Modified SILCS-Pharmb 77.1 9.2 4.4 0.89
Original SILCS-Pharmc 99.6 9.8 4.3 0.88
DOCK 4.0 2.0 1.5 1.2 0.58
AutoDock 4 0 2.1 1.5 0.56
AutoDock Vina 4.2 3.6 2.8 0.73
FPPd 9.3 4.7 3.5 0.81
HSRPd - - - <0.80e

FXa 4F SILCS-Pharma 41.9 11.4 5.1 0.91
Original SILCS-Pharmc 20.7 6.5 3.5 0.81
DOCK 4.0 26.3 4.2 2.6 0.76
AutoDock 4 0.7 0.7 0.7 0.42
AutoDock Vina 3.8 1.5 1.6 0.64
FPPd 12.2–29.7 5.6–6.7 3.5–3.9 0.81–0.87f
HSRPd - - - <0.83e

DHFR 3F SILCS-Pharma 29.3 4.8 3.0 0.79
Original SILCS-Pharmc 0.0 5.5 4.0 0.87
DOCK 4.0 6.3 1.8 1.8 0.66
AutoDock 4 6.3 2.8 2.8 0.76
AutoDock Vina 3.8 1.7 1.2 0.65
FFPd 2.0 1.5 1.3 0.51
HSRPd - - - <0.62e

FGFr1 3F SILCS-Pharma 0.8 1.5 0.8 0.55
DOCK 4.0 1.7 0.8 0.7 0.33
AutoDock 4 0.0 0.6 0.4 0.38
AutoDock Vina 0.0 1.2 1.2 0.50

P38 MAP 3F SILCS-Pharma 6.4 1.8 1.3 0.57
DOCK 4.0 2.0 1.1 0.7 0.39
AutoDock 4 2.0 0.7 0.8 0.45
AutoDock Vina 2.5 2.4 1.8 0.59

ADA 3F SILCS-Pharma 0.0 3.8 3.0 0.80
DOCK 4.0 0.0 4.3 2.7 0.79
AutoDock 4 0.0 1.8 1.3 0.46
AutoDock Vina 0.0 0.0 0.0 0.36

ERantagonist 4F SILCS-Pharma 14.3 10.3 4.6 0.81
DOCK 4.0 0.0 0.8 0.8 0.46
AutoDock 4 6.0 4.2 2.5 0.64
AutoDock Vina 2.8 2.4 1.7 0.59

AmpC 3F SILCS-Pharma 5.0 2.0 3.6 0.70
DOCK 4.0 104.6 9.5 4.7 0.87
AutoDock 4 0.0 1.0 1.4 0.58
AutoDock Vina 0.0 0.0 0.5 0.35
a

The results of the best performing SILCS pharmacophore model based on AUC with a specific number of key features (e.g. 4F means 4 top FGFE ranked features are considered as key features in the model) are shown.

b

The best modified model for HIVPR with the POS feature changed into a HBDON|POS feature.

c

The best results from the original SILCS-Pharm work16.

d

Full Protein Pharmacophore (FPP) and Hydration Site Restricted Pharmacophore (HSRP) results from Lill and coworkers work7.

e

The upper AUC values for HSRP models using different parameters are estimated from figure 7 in Ref.7 as no exact values were reported.

f

Three protein structures were used for FXa in Ref.7, so the range of AUC values is shown here.