[gmx-users] Scaling of Coordinates is Occurring - why?

[David van der Spoel]

chris.neale at utoronto.ca wrote:
>  From what I can find, it seems that you are correct to assume that 
> frozen coordinates should not be scaled: 
> http://lists.gromacs.org/pipermail/gmx-users/2002-July/002054.html and, 
> from the online manual:
> 
> freezegrps:
>     Groups that are to be frozen (i.e. their X, Y, and/or Z position 
> will not be updated; e.g. Lipid SOL). freezedim specifies for which 
> dimension the freezing applies. 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.
> 
> Nevertheless, you say that your frozen coordinates are subject to 
> pressure scaling. I can suggest the following:
> 
> 1. Are you really freezing what you think you are freezing? I see no 
> freeze info in your post so I can't check for myself. i.e. do you have 
> proper freezegrps? did you utilize freezedim? did you include 
> energygrp_excl (which, although required, shouldn't affect frozen atom 
> scaling if you forgot it). Note that your refcoord_scaling only matters 
> if you are using position restraints, which you say you are not.
> 
> 2. Just use position restraints with a large force constant and use that 
> refcoord_scaling = no.
> 
> 3. If you're absolutely sure, file a bugzilla.

This is a known problem. The options freezing and pressure coupling are 
not compatible. Ref_coordscaling refers to position restraints, see 
http://manual.gromacs.org/current/online/mdp_opt.html#pc
There is no general solution for this, since pressure scaling may make 
the box so small that the frozen system is larger than the box (the same 
problem holds for position restraints by the way).

We would like to do this rigorously but as far as I know no-one has 
worked out a solution.

It gets more complicated when constraints are involved.
> 
> -- original message --
> 
> Pressure coupling is not supposed to scale the coordinates of my frozen
> group, yet it is.  My frozen group is a crystal structure that does not 
> have
> any topology assigned to them: no bond, angle, dihedral constants etc.
> specified.  I find that a crystal structure gets compressed.  A sample of
> the pressure coupling section of my .mdp file is given below:
> 
> ; PRESSURE COUPLING
> pcoupl = parrinello-rahman
> pcoupltype = isotropic
> tau_p = 50.0
> compressibility = 0.1 0.1 0.1 0.1 0.1 0.1
> refcoord_scaling = no
> ref_p = 1000.0 1000.0 1000.0 0.0 0.0 0.0
> 
> I'm using Gromacs 4.0.5.
> Any suggestions appreciated,
> 
> 
> 


-- 
David van der Spoel, Ph.D., Professor of Biology
Molec. Biophys. group, Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:	+46184714205. Fax: +4618511755.
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://folding.bmc.uu.se