On 21/05/14 14:57, Remie Fawaz-Touma wrote:
Thank you so much for your replies.
CCP4 refinement file is showing warnings that the Glc units are at
distance (1.4-2.1A) greater than the acceptable one by the program
(1.33A) from each other even though I got the model from a published
pdb file. so the pdb file generated only shows me dotted lines
instead of bonds, but when I do a real space refinement, the Glc come
closer together and show full bonds then. Am I on the right track here?
The problem here, I think, is that the PDB format doesn't allow you to
tell the refinement (or other) program what it needs to know about how
carbohydrates are linked. So, various softwares makes guesses
(estimations) of varying sophistications about the linkages. One
program might not think that two carbohydrate monomers are linked and
pushes them apart via non-bonded contact restraints, whereas another
might think that the monomers should be linked but now seem too far
apart for a reasonable bond.
Not all published pdb files have good geometry for carbohydrate.
The dotted lines (in Coot at least) come from LINK records in the PDB
file. Coot does not use these to determine carbohydrate linkage. If I
remember rightly Coot will draw bond line between monomers if they are
tandem in the coordinates file. It might also try to draw a dotted line
between if the atoms are specified with a LINK.
HTH,
Paul.
