[phenixbb] maximum allowable B-factor value

[Christian Roth]

Hi Pavel,

My resolution is 2.1 Ang.. So its high enough for individual B-Factor 
refinement. The density for this particular atom is not that good, which 
explains the high B-Factor for this. As the other atoms of the ligand have 
reasonble B-Factors compared to the main chain they seems to be nearly fully 
occupied. I am a bit in doubt about the possibilty that between the two 
connected atoms the B-Factor changes so drastically. 
The R-Factors are absolutely O.K  R/Rfree 20,32/25,03 rms bond 0,007 rms angle 
1,363.
I thought already one could maybe allow a bit more deviation for bond and 
angles.

Christian

Am Mittwoch 27 Oktober 2010 18:45:42 schrieb Pavel Afonine:
>   Hi Christian,
> 
> what is the resolution? You may want to tighten up the B-factor
> refinement restraints manually, by making wxu_scale parameter smaller
> (try several values until the refined B-factors "make sense" and the
> R-factors are still ok). Alternatively, you may try group B-factor
> refinement (again, if the resolution is low enough to make individual
> B-factor refinement not favorable).
> 
> Also, the high B-factors may simply reflect the fact that those atoms do
> not have a good density. In this case making the B-factors of bad atoms
> to be similar to the B-factors of good atoms (by tightening the
> restraints) is not a good idea.
> 
> So, as you can see there is a number of nuances that you need to consider.
> 
> Pavel.
> 
> On 10/27/10 11:26 AM, Christian Roth wrote:
> > Dear all,
> >
> > I did a refinement run with a built in ligand. Within the ligand the
> > B-factor jumps between two connected atoms from 60 up ~110, which is i
> > think a bit too much. Is there a way to restrain the B-factors that they
> > could not jump so much. I would accept negative difference density for
> > the atom. Maybe the ligand is a bit degraded.
> >
> > Best regards
> >
> > Christian
> >
> > @Pavel Tanks for your suggestions with occupancy refinement and
> > B-factors.
> 
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb
>