Difference density quality (DDQ): a method to assess the global and local correctness of ­macromolecular crystal structures

Methods for the evaluation of the accuracy of crystal structures of proteins and nucleic acids are of general importance for structure-function studies as well as for biotechnological and biomedical research based upon three-dimensional structures of biomacromolecules. The structure-validation program DDQ (difference-density quality) has been developed to complement existing validation procedures. The DDQ method is based on the information present in a difference electron-density map calculated with the water molecules deliberately omitted from the structure-factor calculation. The quality of a crystal structure is reflected in this difference map by (i) the height of solvent peaks occurring at physical chemically reasonable positions with respect to protein and ligand atoms and (ii) the number and height of positive and negative `shift' peaks next to protein atoms. The higher the solvent peaks and the lower the shift peaks, the better the structure is likely to be. Moreover, extraneous positive density due to an incomplete molecular model is also monitored, since this is another indicator of imperfections in the structure. Automated analysis of these types of features in difference electron densities is used to quantify the local as well as global accuracy of a structure. In the case of proteins, the DDQ structure-validation method is found to be very sensitive to small local errors, to omitted atoms and also to global errors in crystal structure determinations.