Re: Factors that determine the tilt angle of a peptide

Hello Jian--

> We are calculating the structure of a membrane peptide with two
> helices linked with a loop with PISEMA data (chemical shift, dipolar
> coupling) with dihedral angle， hydrogen bond restraints for the
> helical part, we also have some distance restraints for atoms from
> both helices.
>
> From PISEMA experiment we expect to get models with the right tilt
> angle with respect to the membrane normal (i.e., the z-axis), but we
> got models with a variety of tilt angles. For example, we expect to
> have a tilt angle of 12 to 17 degrees, but some of the models have
> tilt angle that is 26, which is way too large. 
> I'm wondering what are the factors that determine the tilt angle of
> the peptide in XPLOR calculation.

Are your restraints well-satisfied? There may be some problem such
that there are large violations causing convergence problems. If your
PISEMA-based restraints are satisfied, then the tilt angle should be
correct. No?

best regards--
Charles

--
Charles Schwieters     email:   Charles <at> Schwieters.org
                       www:     http://schwieters.org/cds

Re: Factors that determine the tilt angle of a peptide

Hello Jian--

> You are right, there are large violations. I'm confused about the
> values that should be used in the CS and DC tables.  Our
> experimental collaborators have given us a table of the CS values
> (let's call it Set 1), but another computational collaborator uses
> the value that is subtracted by the isotropic value, 119.3 ppm for
> non-Glycine residues (let's call this Set 2. In this table, each
> value is the corresponding value from Set 1 minus 119.3).  It seems
> strange for me that when Set 1 was used, the structures look OK, but
> with huge violations for CS; when Set 2 was used, the structure look
> strange but violations are small. What do you think might be the
> problem?

You might check out

http://nmr.cit.nih.gov/xplor-nih/doc/current/python/ref/csa-methods-marvin.pdf

(referenced from)
http://nmr.cit.nih.gov/xplor-nih/doc/current/python/ref/csaPotTools.html

> Do the simulation parameters in refine.py matter much?  Is
> there a script specifically designed for membrane proteins?

please see eginput/assignFit/02refine.py in the Xplor-NIH
distribution for an example.

I hope this helps--
Charles