> I have made a MD simulation with silver atoms ions attached to DNA strands.
> My problem is that the silver ions (positively charged) are sticking
> together very closely, which does not seem physically accurate. I used the
> ff99SB force field and implicit solvent (IGB = 1).
As we have no idea what you are using as a model of silver ions, it is
very difficult to debug/understand. Moreover, implicit solvent with
mobile explicit counterions is very much a cutting-edge research area (and
one I think the AMBER team has *not* been generally trying to tackle).
To simulate mobile ions in implicit solvent is tricky since you need to
account for finite concentration (and impose some boundary so the ions do
not diffuse out infinitely into the vacuum) while at the same time getting
the balance between bulk and direct ion interaction.
If I were to try looking at silver ions, I would first develop a balanced
and simple model for the ions (for example, see Joung/Cheatham, JCTC 2008
as we did for monovalents - noting that higher charge is even more
difficult!) and then apply the models in explicit solvent. Even then, it
will be a difficult project as our current DNA force fields show little or
less than expected salt dependence.
I would also look at a recent paper by Hud/Orozco on TMA and other
interactions with DNA.
--tec3
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Feb 26 2014 - 20:00:03 PST
