[phenixbb] Incorrect solvent modelling in Phenix.refine?

Pavel Afonine pafonine at lbl.gov
Wed Jun 27 16:53:41 PDT 2012

Hi Dave,

could you please send me the data and model files (off-list, directly to 
my email), and tell residue numbers that are close to the residual 
density in question? I will investigate once I have the files.

Most likely it's a footprint of bulk-solvent mask set in regions where 
there is actually no any solvent.


On 6/27/12 1:29 AM, David Briggs wrote:
> Hi all,
> I'm refining a couple of structures of the same protein (well, point
> mutants) and I am consistently seeing large (6sig) negative difference
> map peaks within some cavities within my protein, present in both the
> datasets. As far as I understand, this is because the bulk solvent
> scaling parameters are either incorrect, or being incorrectly applied.
> I have tried a few things like applying the "Refine solvent mask
> (slow)" options, all to no avail.
> Resolution is 2.2Å and 2.5Å. Experimental phasing, Xtriage reports no
> abnormalities in the data.
> Is there anything that I am missing? Do I need to worry about this?
> Regards,
> Dave

More information about the phenixbb mailing list