Hi Chris, from what you described I understand that there are two spices in alternating locations that share the same or nearly the same space. In this case assign one of them altloc A, and the other one altloc B. Then define the constrained group constituting these two species as described here: http://phenix-online.org/documentation/refinement.htm#anch22 If you do so, their occupancies will be refined, and each species will end up with a partial occupancy 0<=q<=1, and the sum of the two group occupancies will add up to one. Also they will not "see" each other through repulsive restraints term and therefore will not be pushed apart. If you have any questions or difficulties to do what I just described above, then please send me the PDB file and point out the atoms in question and explain how you want to refine them, and I will send you back a working example. Pavel On 2/27/12 8:11 PM, Christopher Wanty wrote:

Hi,

I have a problem with a phosphate moving out of position during refinement. The complication is that there is an existing phosphate in the same location as the one I'm trying to place, but the existing one is part of a modified residue, and there is clearly density for the alternative conformation of the phosphate, but it is also clearly not still bound to the residue.

How do I make sure that the phosphate does not move away from where it should be? Attached image show the phosphate more or less where I want it, and you should be able to see where the existing phosphate is attached to the residue. The large red density blob is where phenix.refine tried to put the phosphate last time.

Kind regards, Chris

------------------------------------------------------- Dr. Christopher Wanty [email protected]

Research Associate Biomolecular, Biomedical and Chemical Sciences, Building M310 University of Western Australia 35 Stirling Highway Crawley 6009 Western Australia Australia