1. How does the MR map look like ? Do you see features resembling a real 
solution or just noise ?

2. If the map looks decent, then try running a rigid body refinement using 
Refmac. Does the Rwork/Rfree dro ?

3. At this point you should look at the difference density map in Coot and be 
able to identify "wrong" amino acids e.g. if your MR model had an Ala where 
there is truly a His or Phe, you should see a huge blob of green density 
indicating the missing atoms in your current model.

4. Fix the sequence of your MR model to your true sequence

5. WARNING don't continue to read the next option if you want to learn and have 
fun by yourself !

6. at 1.9 Å ARP/wARP will be able to build most of your structure automatically

Jürgen

P.S. It's likely you have the correct space group simply by gaussian 
distribution over P222 pointgroups, but maybe that's what the Rwork/Rfree are 
indicating  Did you run Phaser in P222 + all other possible space groups or did 
you feed it with P212121 and only ran it in this space group ?
-
Jürgen Bosch
Johns Hopkins Bloomberg School of Public Health
Department of Biochemistry & Molecular Biology
Johns Hopkins Malaria Research Institute
615 North Wolfe Street, W8708
Baltimore, MD 21205
Phone: +1-410-614-4742
Lab:      +1-410-614-4894
Fax:      +1-410-955-3655
http://web.mac.com/bosch_lab/

On Oct 18, 2010, at 7:39 PM, Jyotica Batra wrote:

> Hi All
> 
> I have a dataset at 1.9A, spacegroup-P212121 (unit cell: 37.7, 39.52, 231.72, 
> 90, 90, 90), 
> I used MR phaser and got a structure solution with LLG= 320 (1copy/a.u) .  
> During refinement, the R-free (50%) and R-factors (42%) never go down.
> I have tried refining in phenix and refmac both, but still the high R-free 
> problem persists.  At this point I'm seeking help to know the possible 
> reasons for this high R-free.
> 
> Thanks in advance!
> 
> Jyotica
> 
> 

Reply via email to