The simplest suggestion is to move molecule B by that symmetry operator!
But bewate - the older REFMAC uses the LINK record differently to the PDB - it has now renamed the line as
LINKR and you should check the refmac documentation to see how to set it up. Eleanor Peter Chan wrote:
Good Day, This is my first post on ccp4bb, and I am seeking advice on a problem I have encountered. My crystal structure revealed a nickel ion chelated by 4 nitrogen atoms: 3 from protein chain A and 1 from chain B. A and B form dimers. The nickel was placed in the same asymmetric unit as chain A, and it is linked to the chain B from a different asymmetric unit. The space group is P21 21 21, and from coot, it shows that the symmetry operation required to bring chain B in contact with the nickel ion of its asymmetric unit is: [X+1/2, -Y+1/2, -Z + (0 0 1)] According to http://www.genome.ad.jp/dbget-bin/www_bget?pdb+1UBQ (at REMARK 290) and from my understanding, this corresponds to the code 4556 (from 4555 + the 001 translation). Based on http://www.wwpdb.org/documentation/format23/sect6.html, I believed I could specify this link by including the following into my pdb file: LINK NI NI M 1 NE2 HIS B 103 1555 4556 However, when I run Refmac5 of ccp4i6.0.2, this line appears to be completely ignored by the program. On the other hand, the 3 contacts between Ni and chain A were all detected and refined with no problems, as the previous cycles of refinement have generated such a line in the pdb file: LINKR N HIS A 2 NI NI M 1 HIS-NI Under "Setup Geometric Restraints" in Refmac5, I have attempted to check/uncheck the "Links between symmetry related atoms" option, and tested it in combination with All others if "defined in file only and/or residues are close". So far, I have no luck in getting Refmac5 to recognize this bond. Are there obvious things that I may have missed? Or did I do something completely wrong. Any comments/suggestions would be greatly appreciated. Regards, Peter
