[phenixbb] Glycosidic bonds in cif

[Engin Özkan]

Hi,

I am trying to deposit cif files from phenix.refine using the last 
stable version and the release candidate. In both, I seem many instances of

    C00089 covale NAG D 2012 O4 NAG AM . O4 . BMA D 2013 C1 BMA AN . C1 
. 'Custom glycosidic links need to be generated when the atom names of 
the carbohydrates don't conform to the standard.'

for every glycan-glycan bond, but 'Taken from Monomer Library or GeoStd' 
for ASN-NAG, which does not make sense.

The log file says the ALPHA1-3, ALPHA1-6, BETA1-4 and NAG-ASN links were 
properly applied, but the cif only mentions NAG-ASN. The cif also 
includes a link definition only for NAG-ASN, but none of the other 
glycosidic bonds. The structures appear to have proper geometry for 
glycosidic bonds, so this is probably just a cif-only issue.

Also, is there a way to output the cif where every glycan residue does 
not get a new chain (this one goes up to chain AV=48 chains)? I found at 
least one phenix-refined case in the PDB where one glycan chain gets a 
single chain ID. Admittedly, that might have been fixed post phenix.refine.

Thanks,

Engin

-- 
Engin Özkan, Ph.D.
Associate Professor
Dept of Biochemistry and Molecular Biology
University of Chicago