[gmx-users] hydrophobicity and forcefileds

[Xavier Periole]

Hi Kya.

> I have a membrame active peptide which I have placed in my equilibrated
> mixed bilayer. I know the possible conformations in which it lies (neutron
> diffraction data...from other members of the lab!). The 1st conformation
is
> v. stable, looking at hydrogen bonding with g_hbond. The second
conformation
>
> The 2nd conformation is the peptide lying in the same place except flipped
> over so there is a hydrophobic mismatch ?

Is that a question, an affirmation a result from simulation/experiments ?
Hydrophobic mismatch is not supposed to happen but to be the trigger
of deformation of the peptide(protein)/membrane interface.

> The 2nd one is rather unstable despite me using the same MDP file and more
> or less the same PDB (60 more SPC deleted...ie conflicts). I'm going to
look
> at it with G_confirms...to see the differences between starting structure
> and finishing structure.
>

How do you evaluate the stability of your peptide, system ? If the 60 water
molecules are at the interface between the peptide and the membrane could
be a reason for different comportment of your peptide. Do you use the same
starting points ? I guess not.

> Can forcefields playa part in this un stability????
>

You mean could different force fields yield to different results with the
same
system ? The answer is no in theory but yes in practice. The simulations
are never long enought. But the most important is the starting point.

Can try different starting point : position/orientation of your peptide and
see
if the simulations converge.

Take a look at those papers:
1) Sankararamakrishnan R, Weinstein H.
Molecular dynamics simulations predict a tilted orientation for the helical
region of dynorphin A(1-17) in dimyristoylphosphatidylcholine bilayers.
Biophys J. 2000 Nov;79(5):2331-44.
2) Sankararamakrishnan R, Weinstein H
Positioning and Stabilization of Dynorphin Peptides in Membrane Bilayers:
the Mechanistic Role of Aromatic and Basic Residues Revealed from
Comparative MD Simulations
Ramasubbu Sankararamakrishnan and Harel Weinstein


XAvier