Hi Qiang, although apply_cif_link is probably more elegant way of doing this, you can always use custom bonds (if you can't get apply_cif_link to work): http://phenix-online.org/documentation/refinement.htm#anch86 Using custom bonds you can define a covalent bond between any two atoms. For example, this refinement.geometry_restraints.edits { bond { atom_selection_1 = chain A and resseq 17 and name O atom_selection_2 = chain B and resseq 18 and name N distance_ideal = 1.5 sigma = 0.02 } } will define a bond between these two atoms HETATM 115 O LIG A 17 3.129 18.483 4.947 1.00 44.08 O HETATM 116 N LIG B 18 -2.025 7.355 5.786 1.00 33.96 N You can define any number of such bonds by duplicating bond {} scope of parameters. You can define angle as well. Check .geo file to see if atoms in question are bonded. Pavel. On 4/11/11 2:15 PM, Qiang Chen wrote:
Dear all,
I'm refining a structure which has both N-linked and O-linked glycosylation. I use Phenix to do the refinement. It works well for the N-linked NAG. I defined the link as the following:
apply_cif_link { data_link = "NAG-ASN" residue_selection_1 = "chain A and resname NAG and resid 701" residue_selection_2 = "chain A and resname ASN and resid 518" }
But it doesn't work when I defined the fucose-threonine link as the similar way:
apply_cif_link { data_link = "FUC-THR" residue_selection_1 = "chain A and resname FUC and resid 801" residue_selection_2 = "chain A and resname THR and resid 596" }
The error message is "missing CIF link: data_link_FUC-THR".
Does anyone have the cif link file of FUC-THR?
Thanks a lot!
Qiang