Re: [AMBER] Which is the best thermostat for cMD?

From: David A Case <david.case.rutgers.edu>
Date: Tue, 6 Sep 2016 09:41:37 -0400

On Fri, Sep 02, 2016, Dd H wrote:

> I want to do a MD simulation with a protein in explicit water model. And I
> have to choose a thermostat for temperature regulation. So which is the
> best thermostat for my system? Moreover, which barostat should I use?

I recognize that the manual perhaps makes these things more complex that
they need to be.

The most widely-used thermostat for Amber users, by far, is ntt=3. This
should generally be accompanied by a smallish friction, say gamma_ln=5.
(Some people use even smaller frictions.) Be sure to set ig=-1, or use
some other scheme to prevent re-use of the random number seed on restarts.

Note that all thermostats interrupt Newton's equations of motion: if you are
interested in dyanmical behavior, you really should be running NVE simulations
(generally following NTP equilibration to get a system with the proper
density.) If you are just interested in *qualitative* dynamical behavior,
a Langevin thermostat with small gamma_ln may suffice.

The Monte Carlo barostat is both computationally efficient and samples
the canonical NTP ensemble. The older, "Berendsen" barostat, suppresses
volume fluctuations, but is still often used.

Opinions on this subject vary a lot, and others may wish to pitch in.

...dac


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Sep 06 2016 - 07:00:02 PDT
Custom Search