[gmx-users] pressure coupling for lipid bilayer

[Erik Lindahl]

>
>
>
>You mean  A^2, right !! The difference I observe depending on the size
>of the patch is almost less 4 A^2, for POPC, from 72 to 288 lipids.
>
Of course - nm would have been a little too much :-)

>
>
>Don't get me wrong. I am not saying that parameters are perfect. But
>I think that saying there is no or negligible effect in patch size is going
>way to far for me. The area/lipid is one think, but the order parameter
>is depending on it (cause/consequence whatever) and it's effect on the
>protein membrane function, activity is well recognize now, so I consider
>that the area/lipid has to be well reprocduce, for this kind of simulations.
>

Yes - all the properties like order parameters, fraction of gauche 
bonds, average chain tilt
and even area/lipid can be considered "generalized order parameters" - 
they are extremely
coupled. That's why the it's a good idea to make sure the area is in the 
right ballpark, but
how you get it there is less important :-)

The dangerous part is when some authors argue that they do something to 
get the area/lipid
correct (e.g. apply a surface tension), and since the resulting order 
parameters are good they
consider it a "proof" that you should always use a surface tension, 
while in practice it would
be almost impossible to have the wrong order parameters when the area is 
kept at the correct
value ;-)

Cheers,

Erik