[phenixbb] Fwd: Refinement using new Phenix causes electron density/model mismatch

Nathaniel Echols NEchols at lbl.gov
Mon Jan 12 12:41:25 PST 2009


Hi Miles--

Could you please send me the mtz, sca, and pdb files and the .eff file  
that was written by the GUI?  FYI, phenix.refine_gui and phenix.refine  
should function identically, but until I try running it both ways I  
can't tell whether the problem is in the GUI or the command-line  
version.  There have been changes to phenix.refine since 1.3-final but  
these are independent of the GUI.

Also, could you check the space group/unit cell information extracted  
from the files and confirm that these are correct?  (This should also  
be the same as using the command-line version, but it's the only thing  
I can think of that might cause this problem.)

thanks,
Nat

On Jan 12, 2009, at 12:19 PM, Miles Pufall wrote:

> The bb bounced my email back because the enclosures were too big -  
> let me know if you'd like them.
>
> Thanks -
>
> Miles
>
>
> Begin forwarded message:
>
>> From: Miles Pufall <mpufall at cmp.ucsf.edu>
>> Date: January 12, 2009 12:16:06 PM PST
>> To: phenixbb at phenix-online.org
>> Subject: Refinement using new Phenix causes electron density/model  
>> mismatch
>>
>> Hi Guys -
>>
>> I am having a very puzzling problem using the new Phenix (1.4-3).   
>> I have about 20 structures that I am preparing to deposit with an  
>> accepted manuscript, and decided to do a final refine using the  
>> phenix.refine_gui not only to see if I got better numbers, but to  
>> use the integrated validation tools, which I love.  Unfortunately,  
>> often, but not always, when I refine the pdb against the  
>> *_refine_data.mtz the statistics are a bit worse, and when I open  
>> the resulting *.map_coeffs.mtz in Coot it appears to be shifted  
>> with tons of negative electron density.  Needless to say, this is  
>> making my nervous, so I've tried a few things:
>>
>> 1)  Tried refining the same .pdb and .mtz files in Refmac - I seem  
>> to get the same shifting.
>> 2)  Try using the old .def file left over from the final run (just  
>> to make sure I didn't screw up some setting) - and I get the same  
>> shifting.
>> 3)  Try refining against the original .sca file - and I get the  
>> same shifting.
>> 4)  Try molecular replacement of the refined .pdb against the  
>> original .sca file - and sometimes, but not always, get a shift.
>>
>> I am nervous and concerned - can you help me figure out what  
>> happened?  I am enclosing an example -
>>
>> refine_001.pdb = pdb I refined in new phenix
>> refine_001map_coeffs.mtz = mtz output from new phenix refinement  
>> (lots of red)
>> oste44a_output.sca = original scale file from HKL2000
>> MR.1_refine_data.mtz = data from which I am refining
>> sgk_ost_temp_H_refine_004.pdb
>>
>> Thanks -
>>
>> Miles
>>
>> Miles Pufall
>> Postdoctoral Scholar
>> Yamamoto Lab
>> UC San Francisco
>> Cellular and Molecular Pharmacology
>> Mail Stop 2280
>> 600 16th Street, Genentech Hall S-574
>> San Francisco, California 94158-2517
>> (415)476-4480
>>
>>
>
> Miles Pufall
> Postdoctoral Scholar
> Yamamoto Lab
> UC San Francisco
> Cellular and Molecular Pharmacology
> Mail Stop 2280
> 600 16th Street, Genentech Hall S-574
> San Francisco, California 94158-2517
> (415)476-4480
>
>
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://www.phenix-online.org/mailman/listinfo/phenixbb

-------------------
Nathaniel Echols
Lawrence Berkeley Lab
510-486-5136
NEchols at lbl.gov






More information about the phenixbb mailing list