Skip to main content
. Author manuscript; available in PMC: 2016 Sep 1.
Published in final edited form as: J Comput Aided Mol Des. 2015 Feb 10;29(9):817–836. doi: 10.1007/s10822-015-9833-8

Table 5.

Important validation decisions

Decision Available tools Validation parameter
Does the electron density support the ligand model? Inspect electron density using Twilight, PyMOL, Coot or EDS RSCC < 0.8 suggest significant disagreement with electron density
Is the ligand in a region affected by crystal packing? Inspect symmetry of model using PyMOL or Coot Clashes of symmetry mates with ligand binding region suggest possible perturbations of conformation
Are the B-factors of the ligand significantly higher than the overall average? Inspect B-factors of the ligand using PDB record or Twilight B-factor >200 Å2 corresponds to ~1.6 Å displacement
Is the resolution of the structure sufficiently high to support the model of the ligand? Inspect reflection count and resolution in PDB record Resolutions >2.5 Å likely have a lower data-to-parameter ratio
Is the Kd of the ligand binding high in the crystallization conditions employed? Further biophysical or biochemical studies required Ligand occupancy is likely less than 100 %

Questions that any user of protein–ligand models should ask in order to use the ligand model with confidence