On Wed, 2015-02-18 at 09:39 -0700, Ahmed Ayoub wrote:
> Dear Amber Users,
>
> I have calculated MM/GBSA interaction energy between two proteins and I
> want to translate it into forces or attraction/repulsion and probably a
> generation of torque. I was wondering about a couple of things:
>
> 1) Which terms in the MM/GBSA equation are force generating? My intuition
> says SA term is not and should be excluded. Am I right?
No. They are all force-generating. Remember, the forces are the
negative gradient of the potential energy function. If an energy
contribution is not a constant shift, it contributes to the force.
> 2) If I have a potential energy of interaction of say -30 kcal/mol, how
> much force will be generated between the two proteins?
No idea. The value of the energy itself is not enough to determine
forces. You would either need an analytical form of the binding energy
as a function of distance between the molecules to get an analytical
gradient, or you would need to estimate the forces by some type of
finite-difference method.
> 3) If around the center of mass of the "ligand"-receptor ensemble, the
> interaction energy is unevenly distributed, say 5 kcal/mol on one side and
> 25 kcal/mol on the other side, would that generate any torque? What
> functional form can I use to calculate that torque? Is it the regular
> torque or there is something I'm missing?
Torques are related to forces. The formal definition is t = r x F,
where all quantities are vectors and x is the cross-product. So you can
compute the torque on all of the atoms by computing their forces and
crossing that with their position vector. Keep in mind that the
absolute coordinate system you are using matters here if you are
comparing torques of the same atoms between different conformations.
> Any help will be appreciated.
Binding "forces" from MM/PBSA could be an interesting idea, but I've
never thought about it before and don't even have a grasp of what -- if
anything -- they would really *mean*. But my suggestion is that you
need to go back to the MM/PBSA thermodynamic cycle and start writing out
closed-form equations for the energy terms for whose forces you want to
compute. My naive thought is that they would be some kind of difference
between the forces on the particles in the bound state less the forces
on those same particles in the unbound state (but again, you would need
to demonstrate that the forces are the gradient of the actual potential
energy function you are using).
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Feb 18 2015 - 10:00:02 PST
