[gmx-users] Position restraints on a lipid bilayer

[Mark Abraham]

toma0052 wrote:
> Hello,
>      I am looking to add position restraints to some atoms of a lipid
> bilayer.  There seem to be a lot in the list archives on this subject, but I
> am still a tad confused.  What I would like to do is to position restrain the
> phosphorus atoms in only one leaflet of the bilayer.  So how do I generate
> such a pr file, and where should I include it?  It seems as if things are
> done per molecule, so if I put it in my *.top or *.itp files, the restraint
> is on all P atoms in the system, not just the ones I want.  Is there a way to
> do this so that things are indexed per atom rather than per molecule?

Yep. Make two copies of the lipid molecule with the same parameters and 
different molecule names. Easiest to do this by copying an .itp file and 
changing names and stuff. Then you arrange to use a [ 
position_restraints ] field in one and not the other. If there's not 
already an #ifdef-ed [ position_restraints ] in your .itp file, then use 
an #ifdef like you get normally from pdb2gmx output, and change the 
#ifdef name in one of them so that you can selectively choose position 
restraints in one molecule type or the other. You can share one 
posre.itp file between both these "sorts" of lipid molecules, since the 
atomic index of P will be the same in both molecules. Then set up your 
bilayer such that the molecules in one leaflet precede the other 
leaflet. Now use the [ molecules ] section of your .top with suitable 
naming and ordering to permit yourself the ability to select P atoms in 
the various leaflets for PR. Chapter 5 will have more details on various 
aspects of the above.

Mark