[phenixbb] Low resolution refinement

[Nathaniel Echols]

GMail let me down - the screenshot is actually attached now.

[image: Inline image 1]


On Sat, Aug 23, 2014 at 9:44 PM, Nathaniel Echols <nechols at lbl.gov> wrote:

> On Sat, Aug 23, 2014 at 3:00 PM, Appu kumar <appu.kumar9 at gmail.com> wrote:
>
>>                  I have a data from a crystal which we can process to
>> 4.5A, and we are able to get a good molecular replacement model and it is a
>> tetramer in ASU. I tried refining structure with phenix.refine but Rfree
>> stuck at 0.41. I read a paper published by David Baker and phenix developer
>> in 2013 about the low resolution refinement using a combination of rosseta
>> and phenix. I would appreciate if i get a help on the refinement in rosseta
>> along with phenix. This paper convince me that its worth trying the rosseta
>> and phenix together for low resolution structure refinement.
>>
>
> I think it is worth a try, but I recommend (at the same time) that you
> look for potential explanations for the high R-free, and try other tools as
> well (starting with Xtriage, but also Tom's model morphing program) if you
> haven't already.  The attached screenshot illustrates one example from the
> paper - it is a 4.0Å structure stuck at an R-free of 0.4, with very clear
> errors in the (sequence-idential) model reflected in the difference map.
>  If your map looks like this, that may mean Rosetta et al. will help.
>  Occasionally, it also means that the sequence identity is poor and you
> have to fix this before you and do more refinement.
>
> For a more detailed answer, you should probably email us off-list at "
> help at phenix-online.org" with more details, log files, and, preferably,
> model and data.  As always, these will be kept confidential within Phenix
> developers, but if you are not a P.I., please ask your P.I. if this is okay
> first!  But I will also add some generic advice to the documentation -
> there is a fairly common set of diagnostics and procedures that get
> repeated at workshops or in other interactions with users, but I don't
> think are written down in any one place.
>
> -Nat
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20140823/e20cbfda/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 1isr_1ewt_ref1.png
Type: image/png
Size: 876360 bytes
Desc: not available
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20140823/e20cbfda/attachment-0001.png>