[gmx-users] separating bilayers

[Senthil Kandasamy]

Under certain circumstances, my bilayer leaflets separate.
i.e. instead of the usual

pbc
water
monolayer
monolayer
water
pbc 

I get 

pbc
monolayer
water (thick layer)
monolayer
pbc


I do not know if this is just a virtual error (or does it affect the
simulation?) What causes this flipover? I noticed that it even happens
in some simulations with position restraints on certain lipid atoms.
Does this have something to do with the initial reference coordinate of
the system? e.g. sometimes I set the center to be 0 0 0 (using -center 0
0 0 of editconf rather than the conventional -c option).

Is there a foolproof way to prevent this from happening?

Thanks.

Senthil