[gmx-users] Simulated tempering in gromacs

[yunshi11 .]

Dear List,

I'd like to use "simulated tempering" to increase my sampling efficiency as
I am trying to fold a polymer from its extended conformation. I understand
that GROMACS 5 can easily handle replica-exchange MD, but my system is too
large (>100K atoms) and there would be too many replicas...

Anyway, there seem to be two ways to do simulated tempering, as in the mdp
options:

1) using options of simulated annealing, setting number of annealing
(temperature switching) points, durations after each switch, and
corresponding temperatures;

2) using options of Expanded Ensemble calculations (literally a
"simulated-tempering" option here), setting lower and upper limits of
temperatures, scaling method (to generate temperatures in between), but how
do I set the switch points and durations for each temperature?

The second method seems better, but isn't this "Expanded Ensemble
calculations" function designed for free energy calculations?

Thank you for any advice.

Yun