[gmx-users] position restraint on some solvent molecules

[Anton Feenstra]

Valentin Gogonea wrote:
> Dear GROMACS users,
> 
> Is it possible to impose position restraints on only a selected subset 
> of solvent molecules. If yes, how can this be indicated in the topology 
> file? If this is not possible, can anybody point to me to the function 
> in the GROMACS code which sets up the position restraints for solvent?

It is not very straightforward because position restraints are a property
of the molecule type. That means, that if you want restrained water and
unrestrained water, you will have to make a second 'type' of water molecule,
with the position restraints. You can copy the 'spc.itp' file, give the file
a new name (e.g. spc2.itp) and also fill in a new name (e.g. sol2) for the
[ molecule_type ] in this .itp file. Now, you include this file in your .top
file, and change the 'SOL' line in the [ system ] section into two lines,
one for 'SOL' (the 'default' spc with possible position restraints) and a
second for 'SOL2', which is your new water, without posres. Finally, you
should then be aware that the first 'set' of waters corresponds directly
to the first so many waters in your .gro file, so you may have to change
the order of your water molecules in the .gro file to restrain a certain
set of waters.

> The second question concerns simulation with frozen groups. I was 
> struggling for the past couple of weeks to perform a simulation in which 
> part of a protein and solvent is frozen (using the spc model for water). 
> It happens that after a while a moving water molecule collides with 
> (getting very close to) a frozen water molecule and the simulation 
> crashes giving NaN for coordinates of one water molecule and NaN for 
> energy. I was thinking to impose position restraints for frozen water 
> molecules which are accessible to moving water molecules, but I didn't 
> figure out how this can be done.

If you freeze certain water molecules, it makes no sense to add position
restraints to them also. If you want a 'core' of frozen waters, and a
'shell' of restrained water, see my explanation above for having restrained
and unrestrained water in your simulation. I have my self never used
frozen atoms, but from what I hear they can be problematic in long(er)
simulations and/or in combination with pressure coupling. Beware!


-- 
Groetjes,

Anton
  _____________ _______________________________________________________
|             |                                                       |
|  _   _  ___,| K. Anton Feenstra                                     |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|(   |   )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands   |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610           |
|             | Feenstra at chem.vu.nl - www.chem.vu.nl/~feenstra/       |
|             | "If You See Me Getting High, Knock Me Down"           |
|             | (Red Hot Chili Peppers)                               |
|_____________|_______________________________________________________|