[gmx-users] POPC number

[chris.neale at utoronto.ca]

> I guess that depends on the size of your protein. You should have enough
> lipds such that the protein doesn't interact with its preriodic image.
> 90 POPC seem to be a bit to few for that. In my aquaporin simulations, I
> usually had something like 270 POPE molecules, but if you simulate a
> smaller protein, less lipids may be sufficient. You may also want to use
> a hexagonal box which will allow you have less lipids and water.

A hexagonal box? Great. How does one go about constructing that? Is  
this a cylinder-like box with hexagonal shape or just a triclinic one?  
In the triclinic case, does the PBC membrane turn out as it should?

Thanks,
Chris.