[phenixbb] Defining planarity for Phenix.refine

[Engin Özkan]

I am thoroughly puzzled. I am seeing so many messages in the bulletin 
boards with suggestions of rigid body refinement at lower than 3 
Angstrom resolution datasets, especially with strict arguments of 
data:parameter ratio (which cannot be easily determined due to the 
presence of restraints). Last week, rigid body was suggested for a case 
of 3.3 A structure.

Are there papers out there that advocate residue-by-residue rigid body 
refinement for >3 Angstrom structures?  To me, it sounds like a recipe 
for clashes and bond/angle outliers (when individual coordinate 
refinement can be worked out with proper restraining). Any research to 
back this strategy up, and make it regularly applicable in the 3 to 4 
Angstrom range, or an easy option in phenix.refine? (I have seen grouped 
B-factors tame unruly B-factor refinement, that actually works)

I should say that I have never worked with DNA in crystal structures, 
and due to its structure, it might be better suited to parameterization 
that allows accurate rigid bodies. I just don't know.

I would also like to point out recent work by Axel Brunger's group on 
low-resolution refinement (Schröder et al, Nature, 2010 and references 
therein). Low-resolution refinement, while not straightforward, is 
becoming mainstream.

Engin

On 4/19/10 5:21 AM, Francis E Reyes wrote:
> Peter
>
> Restraining RNA base pairs is a debated topic. Some say that you shouldn't do this and let the X-ray data speak for itself. Some say defining these base pairs should allow the refinement to converge and not distort the rna bases too much.
>
> @ 4A, you're asking for a lot if you're refining with individual_sites. You may want to stay with rigid body refinement with group adp /tls until you've nearly completed the model and then use individual_sites.
>
> While I don't think phenix.refine takes base pairing restraints specifically, one option is to heavily restricting wxc_scale to a small value. Another option is to select your atoms such that A-form helices are not refined with individual_sites and place them precisely with COOT. Another option is to use refinement.geometry_restraints.edits option of phenix.refine. Another option is to switch to CNS in which you can restrain the base pairing in the way you suggest.
>
> Just my $0.02,
>
> F
>
>
> On Apr 19, 2010, at 4:23 AM, Peter Grey wrote:
>
>> Dear all,
>>
>> I have an RNA/Protein big complex at low resolution (roughly 4A). I would like to have the base pairs in the RNA as close as possible to ideal base pairs.
>> I added restraints for distances between base pair hydrogen-bonding atoms but this is not enough to ensure that the bases will be  in the same plane.
>> Could you suggest how to define for  Phenix.refine this planarity ?
>>
>>
>> I am grateful for your advice,
>>
>>
>> Peter
>>
>> _______________________________________________
>> phenixbb mailing list
>> phenixbb at phenix-online.org
>> http://phenix-online.org/mailman/listinfo/phenixbb
> ---------------------------------------------
> Francis Reyes M.Sc.
> 215 UCB
> University of Colorado at Boulder
>
> gpg --keyserver pgp.mit.edu --recv-keys 67BA8D5D
>
> 8AE2 F2F4 90F7 9640 28BC  686F 78FD 6669 67BA 8D5D
>
>
>
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb


-- 
Engin Özkan
Post-doctoral Scholar
Laboratory of K. Christopher Garcia
Howard Hughes Medical Institute
Dept of Molecular and Cellular Physiology
279 Campus Drive, Beckman Center B173
Stanford School of Medicine
Stanford, CA 94305
ph: (650)-498-7111