Re: Using CHARMM-GUI membrane builder with NAMD

From: Bjoern Olausson (
Date: Mon Aug 02 2010 - 06:56:14 CDT

On Monday 02 August 2010 10:43:53 you wrote:
> For the protein, using restraints is not a problem.
> But for water restraints don't make much sense as one would like to let
> water diffuse during equilibration (just keep it away from
> the hydrophobic core). The same is true for lipids. They should be free to
> move in the xy plane of the membrane, but restrained not to go out of the
> plane or turn to much.
> My current system is just an alpha helix in the membrane. Probably it would
> run OK, even If I just heated it without restraints (but I don't have
> enough experience with membrane systems yet to know if this is true or not.
> Or if water would penetrate too deep. Maybe if I heat the wrong way I could
> introduce some artifacts that would take a long time to relax or something.
> This is why I wanted to follow an established membrane equilibration
> protocol.
> So when equilibrating a membrane system is it really necessary to keep
> water out?
> Best regards,
> Ajasja

I would recommend reading about how CHARMM-GUI generates the membrane, then
some of your questions would vanish.

Yes, water restraining makes in most cases no sense.

Why do you want to keep the water out of something where it does not tend to
go voluntary? If at all, you should be concerned, that your water does not
penetrate the membrane headgroups enough so you get a proper hydratisation.
Equilibrations means that your system relaxes and eliminates vacuum regions.
Either by moving structures or diffusing water. Restraints during
equilibration is commonly used to prevent a deformation of your protein (or
whatever Structure) because of to rapid heating, vacuum bubbles, bad
hydratisation, etc.

The system generated by CHARMM-GUI uses equilibrated water boxes in which your
structure is place. All water molecules which heavy atoms are closer then X
ngstrm to your structure are removed. Same (more or less) is done with the
lipids. Again the lipid coordinates are taken from a pre-equilibrated library.
For the area/lipid some initial values are used. All these starting values are
not optimal and in some cases introduce a lot of potential energy. So this is
what you want to overcome in an equilibration and the restraining helps to
keep your structures in shape in case they suffer from effects I mentioned

So just restrain the heavy atoms of your DNA backbone and if you want the
heavy atoms of the lipid head groups (If you really want, put some low forces
on the heavy Atoms of the lipid tails).
In most cases it is not even necessary to slowly heat (even less necessary
when you use restraints and slowly remove them as I suggested in my previous

In fact the CHARMM GUI restraints the water molecules:

define twater sele ( resname TIP3 .and. type OH2 .and. .not. ( segid PWAT .or.
segid WAT* ) ) end
GEO plane -
     xref 0.0 yref 0.0 zref 0.0 zdir 1.0 -
     harmonic outside force @wforce droff 11.0 select twater end

But I am not certain if this is really necessary since I never had problems
with water in the hydrophobic core. Even in self-assembly studies there were
no problems with water in the hydrophobic core.


Bjoern Olausson
Martin-Luther-Universitt Halle-Wittenberg 
Fachbereich Biochemie/Biotechnologie
Kurt-Mothes-Str. 3
06120 Halle/Saale
Phone: +49-345-55-24942

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:54:21 CST