[gmx-users] Re: Energy exclusions for freeze groups, and the pressure

[Andrew DeYoung]

Hi,

I am so sorry to bother you.  If you have time, do you have any ideas or
advice about my trouble below about implementing energy group exclusions to
avoid spurious contribution to the pressure and virial due to frozen atoms?


Thank you so very much!

Andrew DeYoung
Carnegie Mellon University

Hi,

I have a system containing two graphene sheets (with residue names GR1 and
GR2, respectively) plus some liquid.  I would like to hold the two graphene
sheets fixed in space and observe the dynamics of the liquid around it.  

To hold the graphene sheets fixed in space, I used freeze groups:

freezegrps = GR1 GR2
freezedim = Y Y Y Y Y Y ; freeze x, y & z directions

This indeed holds the sheets fixed in space, just as I want.  However, the
pressure increases dramatically, from about 10^3 bar with no frozen atoms to
about 10^29 bar when the graphene sheets are frozen.  I noticed that the
manual says (http://manual.gromacs.org/current/online/mdp_opt.html#neq):

"To avoid spurious contibrutions to the virial and pressure due to large
forces between completely frozen atoms you need to use energy group
exclusions, this also saves computing time. Note that frozen coordinates are
not subject to pressure scaling."

So, it seems that to avoid spurious contribution to the pressure, I need to
exclude interactions between completely frozen atoms.  I used the following
directives in my .mdp file:

energygrps = GR1 GR2
freezegrps = GR1 GR2
freezedim = Y Y Y Y Y Y ; freeze x, y & z directions
energygrp_excl = GR1 GR1  GR2 GR2  GR1 GR2

This series of directives, I think, should tell Gromacs to exclude the
nonbonded interactions between atoms within GR1, between atoms within GR2,
and between atoms in GR1 and in GR2.

However, when I run g_energy to extract the (average) pressure (selecting
"Pressure" from the menu in g_energy), it turns out that the pressure is the
same with or without the energy group exclusion defined by my directive
energygrp_excl above; the average pressure in each case is a whopping
6.91498*10^29 bar (and the RMSDs are the same, too).  So it seems that the
"spurious contribution to the pressure" described in the manual is not
actually being removed by my energy exclusions.

Can you please help me think what I may be doing wrong, or how I can
otheriwse remove the "spurious contribution to the pressure" in the case of
freeze groups?

Or, is the key sentence in the manual actually "Note that frozen coordinates
are not subject to pressure scaling."?  What does it mean that frozen
coordinates are not subject to pressure scaling?  Does this mean that the
pressure is not computed for freeze groups?

Thank you so very much for your time!  I truly appreciate it.

Andrew DeYoung
Carnegie Mellon University