Hi Joe, Ralf correct me if there is a better way of doing this... You can define a bond between any selected atoms: http://www.phenix-online.org/documentation/refinement.htm#anch84 For example: refinement.geometry_restraints.edits { excessive_bond_distance_limit = 10 bond { action = *add delete change atom_selection_1 = chain A and resseq 123 and name SG atom_selection_2 = chain X and resseq 345 and name S1 symmetry_operation = None distance_ideal = 2.1 sigma = 0.02 slack = None } } It is important in your case that you need to define a symmetry operation that relates both atoms in question. There is a trick to get it in Coot but I forgot (Ralf knows - he showed me at some point) - I'm sure it was on phenixbb a few times in the past. After this everything should work. Pavel. On 8/4/10 4:53 AM, Joseph Noel wrote:

Hi Ralf and Pavel,

Unfortunately, Phenix didn't recognize a disulfide bond should exist between the sulfur atoms of DTT and the side chain sulfur atom of the cysteine residue in question. I position everything well in the Fo-Fc and 2Fo-Fc maps and ensured the two sulfurs were approximately 2.1 A apart. However, after refinement the DTT molecule moved so the two sulfurs in question where nearly 3 A away. Maybe this is because it is a mixed disulfide between a DTT molecule and a side-chain Cys rather then what most programs expect which is a disulfide between two Cys residues. Is it possible I need to define the complete chemical entity as a hetatm (the disulfide linked Cys residue? Is there a way to specify that a covalent bond should exist between the SG atom of the Cys in question and the S atoms of DTT with DTT lying across the crystallographic two-fold given its internal 2-fold symmetry?

Joe