[gmx-users] lipid bilayer simulation-simulation setup

[Mohsen Ramezanpour]

Hi Everyone,

Lipid bilayers are (almost all the times) simulated in a rectangular or
cubic boxes where all the angles are 90 between the PBC box sides. They are
usually equilibrated with semi-isotropic Berendsen P-coupling whereas the
production runs use semi-isotropic Parrinello-Rahman p-coupling (this part
make sense). I have two questions:

1) What is the problem if we use a triclinic box with 90 90 120 angles for
bilayer simulation:
If we choose the z-direction as the normal vector to the bilayer. and a
semi-isotropic p-coupling will be applied to the XY plane?


2) There are some publications who used the Berendsen coupling for
production run as well. They do not usually explain why they prefer this
barostat for the production run. It is known that Berendsen barostat is not
producing the target ensemble as good as Parrinello-Rahman.
It might be, however, because of using anisotropic p-coupling where both
the box sides lengths and angles are allowed to change.

Any idea?
Thanks in advance for your input on this questions.

Cheers,
Mohsen

-- 
*Rewards work better than punishment ...*