Hi Nick, could you please try using "custom_nonbonded_symmetry_exclusions" parameter to specify selections for involved atoms to be excluded from non-bonded interactions. If this does not help, please send me PDB file (and optionally data file) and specify which atoms do not behave as expected. Example: custom_nonbonded_symmetry_exclusions="(chain A and resseq 123 and altloc A) or (chain B and resseq 345 and sidechain) or (chain Z and element Zn)" See phenix.refine command-line manual for more examples illustrating atom selection syntax. You can do it from the command line or using the GUI. I will not have email from a few hours from now till mid of next week, so I may not respond very quickly. Pavel On 9/4/12 8:02 AM, Nicholas Keep wrote:

I have an ARG which has nice density for its guanadinium group on a 2 fold. This is the same conformer from the two copies. The occupancy of this chain is set to 0.5 and a second conformer that does not cross the two fold is provided. There is clearly a repulsion term in phenix refines that moves the quanadinium away from the two fold. Do the atoms of the guanadinum group need to have an occupancy of 0.25? Can I specify no repulsion between these symmetry elements?

Best wishes Nick