[phenixbb] Real space refinement

Pavel Afonine pafonine at lbl.gov
Sun Jul 8 13:19:33 PDT 2018

Hi Reza,

> What is the syntax for having phenix.real_space_refine refine a single 
> chain of a multi-chain complex?

unlike phenix.refine, currently phenix.real_space_refine will refine the 
entire content of provided PDB or mmCIF file.

You can edit the file to leave only the part you want to refine and run 
refinement (map does not need any change, use full map). Then paste 
refine part back to the full model. Not ideal and ugly, but may just work.

Alternatively, you may cut out a box with piece of model and map around 
it and do the same as above. This is slightly nicer but still not ideal.

Adding functionality to refine only select part of the model is in to-do 

> I have separated the chains into distinct files and run refinement on 
> a single chain (file) with undesirable results.  It seems that 
> residues adjacent to the density of a neighboring chain get pulled 
> into the neighboring density –possibly because the neighboring density 
> is vacant.  I hope my explanation is clear.

I see. Ok, this is sort of expected. Then use the second option above: 
cut out a box. When cutting out a box make sure you define mask around 
model with some generous buffer and set everything outside to zero.. 
Normally this should do the trick.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20180708/d71cdb83/attachment.htm>

More information about the phenixbb mailing list