[gmx-users] a question about surface tension

[凌未风]

Hi there,

I am doing simulations of a membrane protein,  and I am  trying to apply surface tension to the membrane in order to observe how  the protein respond to it.
I have several questions of how to apply a surface tension to a membrane

1, this is what I am using in my mdp file, could anyone comment on it? I am not sure if I am doing it correctly:

pcoupl                  = berendsen
pcoupltype              = surface-tension
tau_p                   = 5.0
compressibility         = 4.5e-5  4.5e-5
ref_p                   = 200    1.0

Is  my setup correct?  In the manual, it is mentioned that a accurate  enough compressibility for the z direction should be used. What value is  considered as accurate enough?
I noticed that in some papers, people  use "semiisotropic" for option "pcoupltype" and use a negative pressure  (such as - 80 bar) for "ref_p" option. In this way, a surface tension  is applied "in directly". Is th
is recommended? What is the advantage and disadvantage of this method?

2,  It is said that the surface tension converge slowly. Generally for a  regular membrane constituted by POPC, how long does it take for the  surface tension to converge?

I tried so send this message  yesterday, but it seemed that it was not sent successfully. So I paste  it again, sorry for the inconvenience. 



Thanks very much for your help in advance. Any suggestions are highly appreciate!

Cheers,
Ruo-Xu