Re: AMBER: Using TLEAP to set up MD in the crystal lattice space?

From: David A. Case <>
Date: Mon, 13 Aug 2007 18:26:24 -0700

On Mon, Aug 13, 2007, David Cerutti wrote:
> I would like to do an MD simulation in the crystal lattice space. The
> protein crystal is orthorhombic (space group I222), and I have made a PDB
> file containing the protein's symmetry-related neighbors in its unit cell.
> What would be the best way to obtain the topology and starting
> coordinate files to do this simulation in the crystal lattice? Please
> note that although I have reconstructed the crystal unit cell, come
> portions of some of the symmetric units will hang out one side of the
> original unit cells and into neighboring cells.

That should not be a problem.

> I'm also not sure exactly
> how to solvate this system, but I think the best strategy would be to hold
> the protein with tight restraints for successive additions of water
> molecules and simulation in the NVT ensemble, then simulation in the NPT
> ensemble while maintaining the protein atom restraints as small numbers of
> water molecules are added or taken away until the pressure stabilizes at
> 1atm with the right unit cell volume.

You are right that figuring out how many waters to add, and where they should
go, is a challenging task. You need to try to get the starting configuration
as good as possible, since the time required to relax waters from "bad"
positions to good ones might be prohibitively long. You might look at packmol
(, although I have not used it
for this purpose. If anyone on the list has good ideas here, I hope they will
contribute them: this is certainly something I would like to be able to do.
[For example: does anyone know if CHARMM has an approach to this problem?]

The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Wed Aug 15 2007 - 06:07:36 PDT
Custom Search