Re: [AMBER] TI problem

From: David A Case <>
Date: Thu, 11 Oct 2012 21:37:14 -0400

On Thu, Oct 11, 2012, Fabrício Bracht wrote:

> I am trying to run a TI simulation, mutating a zinc bound water
> molecule (H-O-H) into a hydroxyl group (O-H). I have set up the
> simulation by following the instructions on the amber tutorial A9.
> Once I set the two parameter files for the HOH and the OH groups and
> run the mpirun -np $NP sander.MPI -O -ng 2 groupfile I get the
> following error.
> SANDER BOMB in subroutine setup_sc
> All non-softcore atoms must be identical in both systems
> Well, the number of atoms of both systems is exactly the same. But the
> atoms themselves are not exactly equal, since the system with the HOH
> group has "n" charge and the OH has "n-1". Being so, one system has
> "n" sodium atoms and the other has "n-1" (so that the system charge
> remains 0). The number of water molecules in both systems are the same
> and the rest of the protein (non softcore part of the protein) is also
> identical. Knowing this, how am I supposed to make both system's
> non-softcore atoms identical?

Either get rid of the extra sodium ion, or include it in the softcore region.
There is no good reason to require that the charge be 0 in both states.
Remember that you will be doing a parallel transformation for the other side
of a thermodynamic cycle in order to obtain a number that has real physical
reality. As long as you do things the same way in both calculations, things
should cancel.


AMBER mailing list
Received on Thu Oct 11 2012 - 19:00:03 PDT
Custom Search