[gmx-users] PMF from targeted MD vs. Umbrella Sampling/constraints

[anaome]

Dear Gromacs users/developers,

I have calculated the PMF for extending a polymer chain from a looped
perfectly circular structure where both ends are in close contact to a fully
linear slightly stretched conformation. To do this, I used the targeted MD
utility of grompp with the target fully extended structure as input with -rb
and the two end (coarse grained) monomers subjected to a 3D restraint with
k=100. I obtained the PMF by standard integration of the average forces
extracted from the dhdl.xvg files of every of the 50 windows I considered
(init-lambda=0.0,0.02,...,1.00).

I wanted to know exactly to what the force recorded corresponds. Is it the
force along the vectors joining the two particles in the current window to
their counterparts in the target structure? Do these forces correspond to
those that one would obtained with a distance constraint between the two
ends? And more importantly, are the PMFs obtained from this targeted MD
simulation equivalent to those one would obtain from simulations with a
end-to-end distance constraint or umbrella potential?

Thank you in advance for you help
Best wishes

Aymeric

--
View this message in context: http://gromacs.5086.x6.nabble.com/PMF-from-targeted-MD-vs-Umbrella-Sampling-constraints-tp5015591.html
Sent from the GROMACS Users Forum mailing list archive at Nabble.com.