[gmx-users] membrane with OPLS force field

[Volker Wirth]

> standard acyl chain parameters fail when they get too long. I think there
> ref. for that is a Berendsen paper on hexadecane simulations, but I could be
> wrong there. In any event, 'protein' parameters do not appear to
> transfer directly to long acyl chains in a simple way and therefore your
> non-coulombic parameters may also need to be parameterized.
>
> I understand your motivation to do this, and of course your parameters would
> be widely accepted if they produced
> good bulk properties, but we're talking about an entire phd project here, so
> I suggest that you don't plan to
> have this up and running from scratch in the next few months.
>

As I read more about membrane simulation, I found this interesting article:
Shirley W. I. Siu, Robert Vácha, Pavel Jungwirth, and Rainer A. Böckmann
Biomolecular simulations of membranes: Physical properties from
different force fields
J. Chem. Phys. 128, 125103 (2008); DOI:10.1063/1.2897760

They compare GAFF, Berger and Charmm27 simulating a DOPC system with
experimental results. The best ff seems to be GAFF (Generalized Amber
FF, all atom), but there is no 'super' force field for _both_ membrane
and protein.

I am looking for a straightforward method to use in my diploma thesis
(time scale eight months). Combining different force fields
(Berger/GROMOS, Berger/OPLS) for lipids and proteins would be too
complex, also setting up GAFF for another lipid system (DMPC, DMPS and
mixed bilayers) won't be possible and in the moment the idea is to use
the paper's approach to get at least a simulation with **PC lipids.

-- 
###################################
Volker Wirth
Center for Medical Physics and Technology
Biophysics Group
FAU Erlangen-Nuremberg

Email: vwirth%at%biomed.uni-erlangen.de
###################################
(lɯʇɥ˙dılɟ/ɯoɔ˙pɐɟʌǝɹ˙ʍʍʍ//:dʇʇɥ) ¡ʇuǝɹǝɟɟıp ʞuıɥʇ