> Long answer: The amount of solvent may depend on type of simulations. My
> thinking is that you do not want the molecule to interact with its
> periodic image and create artifacts. Electrostatic interactions are the
> longest interactions and they are especially significant in DNA
> molecules. Therefore, the box should be big enough to embody the
> molecule and at least half of the range of electrostatic interactions
> from the surface of the molecule. The box should be big enough in all
> directions because DNA molecule will rotate during long simulations. In
> other words, rectangular box with much different length and height is
> usually not optimal. If you run single stranded DNAs and you start from
> compact structure, consider that the molecule may relax and expand
> significantly.
Remember that in Ewald calcs, the cutoff only applies to L-J
potential, and all atoms see infinite periodic images of each
other.
Bill
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Dec 11 2009 - 17:00:02 PST
