Re: [ccp4bb] Maps look different from auto-mtz vs EDS vs FFT in Coot or CCP4MG.

[Joao Dias]

Dear Kendall,
I would suggest you to run a simulated annealing omit map around the  
region you are interested.

The model bias will be reduced and you will get a more clear answer.

Ciao,
Joao

Joao M. Dias
Ollmann Saphire Lab
The Scripps Research Institute
10550 North Torrey Pines Rd. IMM-2
La Jolla, CA 92037 USA
tel (858)784-8925

On May 24, 2007, at 9:57 AM, Kendall Nettles wrote:


I’d like help in interpreting some mystery density in a structure.

I’m writing a paper about soaking the apo-estrogen receptor with  
different ligands. The apo structure is already released, as pdb  
code 2B23. The question is whether there is a mystery molecule in  
the pocket of the apo receptor. If you superimpose 3ERD, you can  
see where the ligand binds. The problem is that with some maps the  
pocket appears completely empty, and with others, there appears to  
be something there. Protein looks essentially identical with the  
different maps. We have used a few different approaches to identify  
the compound with LC-MS, and are pretty sure there is nothing  
there. For example, we can bind our protein to beads, soak it with  
estradiol, wash extensively,  elute with organic solvent and find a  
great peak for estradiol, but nothing for the apo protein. We have  
also tried non-denaturing MS.


If you look at the 2mFo-DFc map from EDS in Coot or CCP4mg, you see  
mystery density in the pocket. If I  use the MTZ, you see density  
in Coot, but not CCP4MG. I then downloaded the structure factors  
from the PDB and made an MTZ. The map in CCP4MG shows some density,  
but much less that with the map from EDS. When I used FFT to make a  
2mF1-1nF2 map, there is no mystery density in either CCP4MG or coot.


I was ready to submit the manuscript with a picture of the mystery  
density, but now I’m not sure if that is appropriate. Any  
suggestions, as far as how to interpret this mystery density would  
be greatly appreciated.


Best Regards,
Kendall
--
Kendall W. Nettles, PhD
Assistant Professor
Department of Cancer Biology
The Scripps Research Institute
5353 Parkside Dr.
Jupiter Fl 33458

office 561-799-8851
fax 561-799-8805
cell 561-306-7566

[ccp4bb] Maps look different from auto-mtz vs EDS vs FFT in Coot or CCP4MG.

[Kendall Nettles]

I¹d like help in interpreting some mystery density in a structure.

I¹m writing a paper about soaking the apo-estrogen receptor with different
ligands. The apo structure is already released, as pdb code 2B23. The
question is whether there is a mystery molecule in the pocket of the apo
receptor. If you superimpose 3ERD, you can see where the ligand binds. The
problem is that with some maps the pocket appears completely empty, and with
others, there appears to be something there. Protein looks essentially
identical with the different maps. We have used a few different approaches
to identify the compound with LC-MS, and are pretty sure there is nothing
there. For example, we can bind our protein to beads, soak it with
estradiol, wash extensively,  elute with organic solvent and find a great
peak for estradiol, but nothing for the apo protein. We have also tried
non-denaturing MS. 

If you look at the 2mFo-DFc map from EDS in Coot or CCP4mg, you see mystery
density in the pocket. If I  use the MTZ, you see density in Coot, but not
CCP4MG. I then downloaded the structure factors from the PDB and made an
MTZ. The map in CCP4MG shows some density, but much less that with the map
from EDS. When I used FFT to make a 2mF1-1nF2 map, there is no mystery
density in either CCP4MG or coot.

I was ready to submit the manuscript with a picture of the mystery density,
but now I¹m not sure if that is appropriate. Any suggestions, as far as how
to interpret this mystery density would be greatly appreciated.

Best Regards,
Kendall
-- 
Kendall W. Nettles, PhD
Assistant Professor
Department of Cancer Biology
The Scripps Research Institute
5353 Parkside Dr. 
Jupiter Fl 33458

office 561-799-8851
fax 561-799-8805
cell 561-306-7566