[Fwd: Re: [gmx-users] best structure]

[David van der Spoel]

-------- Forwarded Message --------
From: UCT Staff Member - Jackson <Jackson at SCIENCE.uct.ac.za>
Reply-To: Jackson at SCIENCE.uct.ac.za
To: David van der Spoel <spoel at xray.bmc.uu.se>
Subject: Re: [gmx-users] best structure
Date: Fri, 11 Feb 2005 08:31:26 +0200
Dear David
Thanks for the reply.  

I am looking at an insect neuropeptide which binds to a G-coupled
protein transmembrane receptor.  I have collected nmr data in water/dpc
micelle solution to try and mimic the membrane surface.  I am using the
nmr data to restrain the peptide during the dynamic runs.

We I built the peptide and minimized it I got 2 phenylalanine sidechains
on the same surface of the peptide even though they are not next to each
other.  This made sense because they could then project into the
hydorphobic membrane.  However, in looking at the structure in more
detail I found that two of the peptide bonds, cys and gly, were cis.  I
therefore rebuilt the molecule with all trans peptide bonds and again
minimized.  The new structure was higher in energy and had the
phenylalanine side chains on oposite sides of the peptide.  I therefore
did not know which starting structure to use in my dynamic simulation. 
I am using a biphasic water/decane box for this.

To make matters worse my original construct which was estheically
pleasing was accedental and so I do not know if there are other grossly
different starting conformations.  I have now built several others
varying which peptide bonds are cis.  However with 10 aminoacids this
leads to a large number of possibilities.  

To try and get arround this I did vacuum dynamics at 600K at which
temperature I did get some cis/trans isomerization.  From this I ended
up with 3 low energy conformers, (1) all trans,(2) one cis bond and (3)
two cis bonds.

These I used as starting structures in my biphasic box simulation. 
After cluster analysis I end up with 4 clusters.  Only one of the
dynamic runs gave 2 clusters with significant populations while the
other two runs gave only one cluster each.

My problem now is to compare these clusters.  AT the moment I am looking
at the average noe violations.  I am also looking at the total
peptide-peptide, peptide-water and peptide-decane energies to see if
this gives a clue.

It is unlikely that in nature you would get any cis/trans isomerization
but it is possible that during the synthesis a cis isomer is produced.

Can you make any suggestions?

Thanks
Graham

 Looking at the most populated structure is fine for each dynamic run
but I

David van der Spoel wrote:
> 
> On Wed, 2005-02-09 at 09:08 +0200, UCT Staff Member - Jackson wrote:
> > Dear Users
> > I have done a restrained molecular dynamics of a cyclic peptide in a
> > biphasic solvent box.  I have used different starting conformations and
> > collected 100 structures during each run.  After cluster analysis I get
> > 3 clusters.  How do I tell which is the 'best' structure.  If I was
> > working in vacuum I could use the potential energy as a criterion but I
> > am working in solution.  I have tried using restraint violations, both
> > in terms of the number of violations and their magnitude.
> If you've sampled sufficiently, then the most populated cluster is the
> best one. Of course it can be tricky with cis/trans isomerization.
> 
> >
> > In theory I could start the simulation with all cis peptide bonds.  The
> > energy would be very high but at 300K I would not get cis-trans
> > isomerization.  Hence I could end up with energetically unrealistic
> > structures as long as they did not violate my restraints.
> >
> > In essence the difficulty is to compare the energy of different
> > simulations on different conformations of the same molecule but in a
> > solvent box of slightly different size and number of solvent molecules.
> >
> > Any help
> >
> > Grahan
> --
> David.
> ________________________________________________________________________
> David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group,
> Dept. of Cell and Molecular Biology, Uppsala University.
> Husargatan 3, Box 596,          75124 Uppsala, Sweden
> phone:  46 18 471 4205          fax: 46 18 511 755
> spoel at xray.bmc.uu.se    spoel at gromacs.org   http://xray.bmc.uu.se/~spoel
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> 
> --
> David.
> ________________________________________________________________________
> David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group,
> Dept. of Cell and Molecular Biology, Uppsala University.
> Husargatan 3, Box 596,          75124 Uppsala, Sweden
> phone:  46 18 471 4205          fax: 46 18 511 755
> spoel at xray.bmc.uu.se    spoel at gromacs.org   http://xray.bmc.uu.se/~spoel
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

-- 
David.
________________________________________________________________________
David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group,
Dept. of Cell and Molecular Biology, Uppsala University.
Husargatan 3, Box 596,          75124 Uppsala, Sweden
phone:  46 18 471 4205          fax: 46 18 511 755
spoel at xray.bmc.uu.se    spoel at gromacs.org   http://xray.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++