 |
 |
I am convinced that there is a continuum. There are tightly bound waters with high occupancies and low B-factors, and loosely bound waters with low occupancies and (very) high B-factors. I cannot believe that a water with a B-factor of 79 Å^2 is bound and when heated to 81 Å^2 it suddenly flies away from the protein. It therefore boils down to a subjective decision: At which B-factor do I reject a water molecule? I use a cutoff of 85 Å^2, but this is purely pragmatic. Occupancy refinement of water molecules does not give good results in my hands and I find that for something as small as a single water molecule, the B-factor takes care of the occupancy as well. If the sulfate density looks sulfate-like, and fitting a water molecule would lead to difference density peaks and unrealistic low B-factors, I would leave the sulfate in. Again, this is a subjective decision. Best, Herman -----Ursprüngliche Nachricht----- Von: CCP4 bulletin board [mailto:[log in to unmask]] Im Auftrag von Mohamed Noor Gesendet: Dienstag, 23. August 2016 15:00 An: [log in to unmask] Betreff: Re: [ccp4bb] Ion placement On a similar note, what should I do with water molecules with B-factor > 80 A^2? Should I refine their occupancies as well, or model something else? I tried to Google why 80 A^2 is used in Coot and Phenix as default values, but I cannot find any reference. Is there any theoretical basis for choosing this value? Is there any link with the Wilson B? Mohamed On Tue, 23 Aug 2016 11:48:21 +0100, Mohamed Noor <[log in to unmask]> wrote: >Dear all > >I have placed a few non-metal ions in my model, such as SO4(2-) and PO4(3-). After anisotropic B-factor refinement, the B-factor of these ions are about 140 A^2 and the residues closest to them are about 50 A^2. The model is refined to 1.9 A, with an average B-factor of 60.4 A^2. > >Although the local CC is at least 0.7, does the high B-factor mean these should be replaced with something else? If so, should it be something lighter? > >Thanks. >Mohamed