On Sat, Mar 25, 2017 at 3:05 PM, Ramin Salimi <ramin.salimi01.utrgv.edu>
wrote:
> Adrian
>
> That was very informative answer. Thanks very much.
>
> However, I do have questions as well: They should have used nmropt not
> jar. There seems to be nothing as relaxation period within the nmropt
> codes. It seems to me that they wrote a loop script to pull the DNA for 1
> A0, do the relaxation period; then stretch it one more unit, and it goes on
> to reach 110 A0. Am I right?
>
> Secondly, Column 4 being the work but I said the derivative of column 4 vs
> column 1 should give the force vs extension was due to the relation
>
> Work Done=Force*Displacement
>
> I thought that the force coming from the potential energy should be the
> same force in the above relation. Is that right?
>
It's unclear what you mean by "the force coming from the potential
energy". The Work that Adrian is referring to is the work done by the
restraining potential as it moves. This is a non-equilibrium,
path-dependent quantity (i.e., it is not a state function and will vary --
often by large amounts -- from one simulation to the next.
You *can* get back to state variable quantities from non-equilibrium work,
but only if you run ensembles of simulations and employ relations like the
Jarzynski equation (which is why the variable in Amber is named "jar").
So the take-home message here is that none of the energies you calculate
from a single Steered MD simulation is meaningful -- only when combined
with many SMD simulations via the Jarzynski equation do the reported works
become meaningful.
HTH,
Jason
--
Jason M. Swails
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Mar 26 2017 - 17:30:02 PDT
