[gmx-users] membrane protein simulation

[Justin Lemkul]

On 4/22/15 4:07 PM, Mostafa Javaheri wrote:
> Dear Justin
> I am going to simulate a homo trimer trans-membrane protein; Base on the
> crystallographic structures there is 7 phosphatidyl glycerol phosphate
> (PGP) per each monomer and also 3 glycolipid molecules (S-TGA-1, or 3-HSO 3
> -Galpβ1-6ManpR1-2GlcpR-1-archeol) located inside the trimer on the
> extracellular side of membrane. In the membrane protein tutorial of gromacs
> 1,2-dipalmitoyl-*sn*-glycero-3-phosphatidylcholine (DPPC) is introduced as
> the standard lipids, so should I treat PGPs and glycolipids as ligands and
> going through protein ligand complex tutorial?
> Or treat glycolipids as ligands, continue the membrane protein tutorial and
> use DPPCs instead of PGPs?
> Is it OK if I replace PGP with DPPC from the standpoint of simulation?

The tutorial is a simple example.  It is absolutely NOT any sort of "standard 
lipid" or model for what you should do.  It's a demonstration to teach you some 
skills and a manner of thinking about a problem.

Choosing a proper force field is issue #1.  The Berger parameters are OK, but 
there are more recent parameter sets that are vastly better.  Issue #2 is what 
lipid(s) is(are) relevant.  In your case, you've got something very different 
that needs to be modeled accordingly.  You need to simulate what's functionally 
relevant, not what a simple tutorial example provides.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================