[phenixbb] refinement questions

Pavel Afonine PAfonine at lbl.gov
Fri Aug 10 09:33:08 PDT 2007


Hi,

first of all I would suggest to use the latest available PHENIX. This 
will make sure you get rmsd and not mean values for the deviations from 
ideal stereo-chemistry. CCI Apps is out of date right now, so the best 
if you use phenix.refine from the latest PHENIX installer (1.3 beta 
pre-release version).

You can look at the NEW updated phenix.refine documentation here:
http://www.phenix-online.org/new-site/documentation/refinement.htm
It has more examples and general information than previous version. In 
particular, it has the answers to all questions that you are asking below.

At 2.5A resolution, I wouldn't worry too much about these small values 
that you get (supposing we look at rmsd): large deviations from ideal 
values at resolutions close to low (like 2.5A) are hardly justified.

Please let me know if you have any more questions!
Pavel.

Sudharsan Sridharan wrote:
> Hi all,
>
> I'm starting to refine a structure using phenix.refine. Almost all my
> model residues have been built starting from a Sharp map. I've been
> trying phenix.refine with and without SA and getting bond angle rmsds
> of 1.0 -1.1 (2.5 A data, using mlhl targets; example below) (yet to
> add water and also trying water picking in phenix.refine). I'm
> wondering if this sounds low (I have read discussions on bond and
> angle rmsds). More importantly I'd like to know how to work with
> different parameters (weights etc) during refinement in phenix. Could
> you please point to what params are safe to manipulate.
>
> Thanks,
> Sid.
>
>
> REMARK ******************** REFINEMENT SUMMARY: QUICK FACTS *******************
> REMARK Start: r_work = 0.3637 r_free = 0.4003 bonds = 0.010 angles = 1.175
> REMARK Final: r_work = 0.2607 r_free = 0.3236 bonds = 0.007 angles = 1.017
> REMARK ************************************************************************
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://www.phenix-online.org/mailman/listinfo/phenixbb
>   



More information about the phenixbb mailing list