Dear Kelly, The way I proceed in these cases is to make a "real" alternative conformation with the same chain ID and residue number. Example: HETATM 1260 N1 AEPE L 1 0.875 15.892 20.141 0.50 7.15 N HETATM 1261 N1 BEPE L 1 0.885 15.884 20.143 0.50 43.64 N HETATM 1262 C2 AEPE L 1 1.461 16.363 21.411 0.50 7.56 C HETATM 1263 C2 BEPE L 1 1.463 16.379 21.413 0.50 43.79 C HETATM 1264 C3 AEPE L 1 0.517 15.990 22.572 0.50 7.83 C HETATM 1265 C3 BEPE L 1 0.544 16.014 22.575 0.50 43.83 C
If you used two "separate" ligand molecules, this is the first thing to try to solve the issue with the bonds. To avoid this kind of issues, I make alternatives for all atoms of the ligand. I do not have experiences with phenix, but in my hands in refmac there are still residual repulsions between the alternate conformations, which as far as I know have not yet been resolved. For this reason I switched to Buster where I do not observe these residual repulsions. Best regards, Herman -----Original Message----- From: CCP4 bulletin board [mailto:ccp...@jiscmail.ac.uk] On Behalf Of Kelly Daughtry Sent: Wednesday, January 20, 2010 2:33 PM To: CCP4BB@JISCMAIL.AC.UK Subject: [ccp4bb] ligand with alternate conformation? Hello all, I am refining a structure (using REFMAC currently, but have tried with phenix as well) with a ligand with alternate conformations. I have added each conformation to the pdb file with 0.5 occupancy and named as A and B and have been able to refine that. The problem is only two atoms are in an alternate conformation, while I have 17 atoms total in my ligand (which are positioned the same - i.e. a carboxyl and sugar ring I only see density for one conformation). After a round of refinement with A and B conformations of the entire ligand, I have two conformations of my ligand (one shifted relative to the other in the density) in which the second conformation does not fit into the density, except for the 2 atoms of concern. If I refine with just the two atoms that are different labelled as partial occupancy A and B, a bond is broken in my main (or rather 'productive') ligand binding conformation, and the bond remains with the 2nd conformation. My question is should I make a LINK in the cif file for this ligand to restrain the bond I want to keep? Is that the best way to go about this? I initially tried putting both conformations of the two atoms into the CIF file, but that is not correct as the CIF file thinks that they are both fully there (i.e. no occupancy in the CIF file) (used phenix.elbow to generate CIF files). Any suggestions on how to move forward? Thanks in advance, Kelly ******************************************************* Kelly Daughtry PhD Candidate Department of Physiology and Biophysics Boston University School of Medicine 590 Commonwealth Ave R 390 Boston MA, 02215 (P) 617-358-5548 *******************************************************
