[gmx-users] Membrane Generating Algorithms and Tools

[Xavier Periole]

on 11/5/02 3:55 PM, Erik Lindahl at lindahl at stanford.edu wrote:

> Tivadar Orban wrote:
> 
>> Hi,
>> I would like to build a lipid bilayer (POPC/POPS=4:1).
>> Did somebody used the algorithm provided by Ilya Balabin at the site:
>> 
>> http://www.ks.uiuc.edu/~ilya/Membranes/
>> 
>> Or you suggest just using a model assembled by a long term MD
>> simulations on a system composed by water and lipid units in the given
>> ratio?
> 
> I would either
> 
> 1) Start with a pure POPC bilayer and replace 1 in 4 lipids.
> or
> 2) Simulate water+lipids from scratch - we've "folded" membranes in
> 30-100 ns, and you don't need very long cutoffs just to fold it. Once it is
> in a membrane form you can increase the cutoffs and/or use PME.
> 
> Cheers,
> 
> Erik

I would do number 1) from Erik's. Equilibration of lipid is
quite long so having a part of it already equilibrated is
probably faster, even though it'll have do equilibrate
again in its new environment.

After a rapid look at the link you gave, the method he proposes
seems to be nice and probably works well. There are different
methods, hopefully equivalent at the end.

good luck
Xavier

-- 
Xavier Periole
Physio & biophys Dept.
Mount Sinai Hospital
New York, NY