Hi All,

the file ending with ".geo" has a complete footprint of all the geometry restraints used in phenix.refine run. For example, if you want to check that atom A is bonded to atom B with given bond length, bond angle, corresponding sigmas, and so on, just find the corresponding record in ".geo" file. If you don't see it there, then it means this particular bond (angle, ...) was not restrained.

Pavel.

On 10/16/2008 3:16 PM, Engin Ozkan wrote:
I don't have access to phenix, ccp4, or any of my files right now (our 
systems are being upgraded), but I'll make some guesses that may help 
you, I might be off the mark.
I think the data_link determines the linkage type, so you should have 
ALPHA1-4 and BETA1-4 for different mannoses in your parameter file.
You can also call alpha and beta mannoses different names, I think MAN 
is usually alpha mannose, and BMA is beta, and give phenix different cif 
files to phenix.

I suggest you look into the cif descriptions, runs a few phenix runs, 
and convince yourself that everything checks out chemically.

Engin

Patrick Loll wrote:
  
OK, so I think you are implying that if I feed phenix.refine 
the MAN-a-D.cif, it will assume that any MAN residue it encounters in 
the PDB file is alpha. Is that right?

Does this mean that if you have examples of both alpha and beta 
mannose residues in a structure, you would need to give one of them a 
different residue name?

Thanks,

Pat

On 16 Oct 2008, at 4:52 PM, Engin Ozkan wrote:

    
Hi Patrick,

Let me give my best shot, I don't have my files with me. There might be 
some mistakes.

You can give phenix.refine the .cif file on command line for 
MAN-a-D.cif, which you can get from ccp4 sketcher (and I think under 
phenix, somewhere close to where you found mon_lib_list.cif).  Then you 
should describe the linkage as an alpha linkage by
data_link = ALPHA1-4  if you have 1-4 linkage, by something like this. 
In the mon_lib_list.cif, I believe they have all recognized linkage 
types somewhere down the file, where you can see the proper description 
for ALPHA1-4, ALPHA1-6, BETA1-4, and other common sugar linkages.

refinement.pdb_interpretation.apply_cif_link {
  data_link = ALPHA1-4
  residue_selection_1 = chain X and resname MAN and resid 905
  residue_selection_2 = chain X and resname MAN and resid 906
}

I also delete hydrogens and the removed O1 in coot before I do 
refinement in phenix, but I am not sure if this is necessary (Phenix 
supposedly knows the deletion, it is in the description of the bond).

Good luck,

Engin

Patrick Loll wrote:
      
Hi all, 

I'm starting to use phenix.refine and have some questions about 
defining modifications during refinement:

I want to refine a structure that includes several alpha-D-mannose 
residues.

I see that  mon_lib_list.cif contains, under the "LIST OF 
RECOGNIZED MODIFIED MONOMERS," the following line:

MAN-a-D   MAN-b-D  SUG-a-D  MAN   alpha_D_mannose            D-pyranose

Question:

Where/how do I invoke this modification? I.e., how do I tell the 
refinement program that a particular sugar is an alpha, and not a 
beta, mannose?

Clearly I can't label the residue MAN-a-D in the PDB file, as this 
would step on the chain identifier.

I'm guessing that I label each sugar MAN in the PDB file, and then 
insert some instructions about applying the modification in the CIF 
file that I supply to the refinement program. However, I'm not able to 
find an example of the appropriate syntax. 


Thanks,

Pat Loll

---------------------------------------------------------------------------------------

Patrick J. Loll, Ph. D.                  

Professor of Biochemistry & Molecular Biology

Director, Biochemistry Graduate Program

Drexel University College of Medicine

Room 10-102 New College Building

245 N. 15th St., Mailstop 497

Philadelphia, PA  19102-1192  USA


(215) 762-7706

[email protected] <mailto:[email protected]>


------------------------------------------------------------------------

_______________________________________________
phenixbb mailing list
[email protected] <mailto:[email protected]>
http://www.phenix-online.org/mailman/listinfo/phenixbb

        
_______________________________________________
phenixbb mailing list
[email protected] <mailto:[email protected]>
http://www.phenix-online.org/mailman/listinfo/phenixbb
      
---------------------------------------------------------------------------------------

Patrick J. Loll, Ph. D.                  

Professor of Biochemistry & Molecular Biology

Director, Biochemistry Graduate Program

Drexel University College of Medicine

Room 10-102 New College Building

245 N. 15th St., Mailstop 497

Philadelphia, PA  19102-1192  USA


(215) 762-7706

[email protected] <mailto:[email protected]>


------------------------------------------------------------------------

_______________________________________________
phenixbb mailing list
[email protected]
http://www.phenix-online.org/mailman/listinfo/phenixbb
  
    

_______________________________________________
phenixbb mailing list
[email protected]
http://www.phenix-online.org/mailman/listinfo/phenixbb