On Wed, Sep 21, 2016 at 4:11 PM, Chitrak Gupta <chgupta.mix.wvu.edu> wrote:
> Dr. Simmerling,
>
> Thank you for the response. I should paraphrase my question to this. What
> is the best way to fix concentration of peptides (multiple molecules) in an
> implicit solvent simulation? Is it possible to define flat-bottom
> positional restraints?
>
It is non-trivial, I believe. Using restraints or constraints will
introduce bias into the simulation that will have (I believe provably)
different thermodynamic distributions than the "desired" distribution (how
much is unknown). Implicit solvent simulations without boundary conditions
are, by definition, infinitely dilute. I believe there is literature out
there on using restraints to impose a concentration (and therefore a
"standard state") on ligands in receptor-ligand binding calculations, but I
think there is some complexity to backing out the desired thermodynamics
and defining the proper restraints.
Keep in mind you'll also be introducing additional degrees of freedom,
which would impact quantities like heat capacities and molar enthalpy
(which is far less dependent on concentration than entropy or free
energies).
By far the best way to fix concentration is to use periodic boundary
conditions (OpenMM supports GB with PBC, if that's something you wanted to
try). But restricting yourself to Amber, I would suggest going to explicit
solvent or prepare to do some intensive lit searches for how people have
done this in the past (particularly the recent past).
HTH,
Jason
--
Jason M. Swails
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Sep 21 2016 - 18:00:04 PDT
