[phenixbb] Building DNA

Nathaniel Echols nechols at lbl.gov
Fri May 14 14:53:48 PDT 2010


On Fri, May 14, 2010 at 2:41 PM, Rajagopalan, Senapathy <
SRajagopalan at tmhs.org> wrote:

> I am solving a protein-dna complex and Autobuild is able to build both the
> protein as well as DNA chain (one at a time). However, when I try to do the
> refinement, I get an error saying “Phenix could not find the geometry
> restraints in the standard monomer library for one or more residues in the
> pdb file... DG: 6 copies, DA: 3 copies, DT: 43 copies”. It seems like
> Autobuid uses the nomenclature DA, DC, DG and DT for the DNA bases while
> Phenix.Refine either doesn’t like the nomenclature or the restraint files
> uses a different one. It still however lets me do the refinement...although
> I don’t know if it just ignores the DNA part. Can someone enlighten me on
> this.
>

This is just the GUI giving you the error, right?  I think this is just a
stupid case-sensitivity error I made - should be an easy fix.  Could you
please send me (not the list!) a relevant part of the PDB file (doesn't need
to be the entire thing, just enough to give this error) so I can debug the
problem?

If phenix.refine itself has any problem finding the appropriate restraints,
it will quit with an error almost immediately after you start it - if this
doesn't happen, you should be okay.

Also, the nomenclature Coot seems to like is Ad, Cd, Gd, Td for the DNA and
> doesn’t like the above phenix style. In order to manipulate in coot, I had
> to rename all the bases...is it possible to have a format that is compatible
> both in Phenix and Coot or am I missing something obvious here.
>

I really hope this isn't necessary - we should be using a similar monomer
library.  What happens when you try to refine the file that was okay in
Coot?  Could you send me that too (or an appropriate excerpt)?

thanks,
Nat
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20100514/20b26c00/attachment-0003.htm>


More information about the phenixbb mailing list