Skip to main content
. 2012 Mar 1;7(3):e32036. doi: 10.1371/journal.pone.0032036

Figure 2. Two examples from the data set used to validate the WaterDock method.

Figure 2

Yellow spheres: predicted water sites, red spheres: water molecules observed in at least two experimental structures, blue spheres: water molecules observed in only one experimental structure. HIV-1 protease bound to the inhibitor KNI-272 (A). All 9 consensus water molecules and all 6 non-consensus water molecules are correctly identified. One non-consensus water molecule is in between two predictions, resulting in a false positive. This water molecule was resolved only in 3FX5 with a temperature factor of 42 Å2, so the over-prediction may be due to the uncertainty in the water molecule's position. GluR2 ligand binding core bound to AMPA (B). All water molecules within the binding site are correctly predicted.