Hi,
This is a common problem in NMR structures. NOEs simply don't offer
nearly enough information to define the structure by themselves,
and the repulsive van der Waals model used by xplor isn't terribly
accurate. (If you want to see what Ramachandran regions are allowed
by the xplor vdw potential, you could take a look at my very first
paper, which appeared way back in 1991.)
There are two easy things you could do to improve your structure:
1. Use Gabriel Cornilescu's TALOS to obtain a bunch of backbone
torsion angle restraints. TALOS only requires your chemical shift table
as input, and its predictions are typically 97-98% correct.
To get TALOS, you'll need to talk to Frank Delaglio.
2. Use my DELPHIC torsion potential (better known by xplorers as the
'rama' potential) in your script. The DELPHIC torsion potential
seeks to
move the structure into commonly-seen combinations of torsion
angles. If your structure is underconstrained, it won't do much to
improve its
precision. But if your NOEs are fairly complete, it will improve
your structure's accuracy a bit, and will dramatically improve its
Ramachandran plot.
The DELPHIC torsions work on both backbone and sidechain torsion
angles, and are available for both proteins and nucleic acids.
To use the DELPHIC torsions, take a look at
eginput/sry/sry_final.inp (if you prefer the old-style xplor
scripts)
eginput/gb1_rdc/anneal.py (if you prefer python), or
eginput/marvin/cvn/sa_pass2.tcl (if you prefer tcl)
Hope this helps. Let me know if you run into difficulty.
--JK
On May 4, 2007, at 11:37 AM, kunal bakshi wrote:
> Hello everyone,
> I am trying to generate a peptide
> structure based on a set of NOE distance restraints using 'Sa.inp'.
> After refining the structure I get a set of acceptable structures
> with RMSDs between 0.5-0.7 but when I try to generate a
> Ramachandran plot most of them(60-70%) lie in unacceptable regions.
> I have checked my NOEs thoroughly and am quite confident of them.
> could anyone help me see where I am going wrong
> thanks
> Kunal Bakshi
>
> Ahhh...imagining that irresistible "new car" smell?
> Check out new cars at Yahoo! Autos.
> _______________________________________________
> Xplor-nih mailing list
> Xplor-nih at nmr.cit.nih.gov
> http://dcb.cit.nih.gov/mailman/listinfo/xplor-nih
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://dcb.cit.nih.gov/pipermail/xplor-nih/attachments/20070504/93f511c6/attachment.html
