AMBER Archive (2007)

Subject: Re: AMBER: ntt parameter in vacuum

From: David A. Case (case_at_scripps.edu)
Date: Mon Sep 03 2007 - 10:51:53 CDT

On Mon, Sep 03, 2007, Age.Skjevik_at_student.uib.no wrote:
>
> ....learning recently that the Langevin
> thermostat is not suitable for MD Production when using expicit solvent.

I think you are reading more into these emails than is warranted. The
Langevin thermostat is probably the best one available in Amber if you wish to
sample a canonical distribution, such as in estimating thermodynamic
quantitites.

For time-dependent properties, this and all other thermostats have some
artifacts, although these are probably fairly mild for default values of
parameters like gamma_ln. This is quite different from saying that the
Langevin procedure is "not suitable": there are trade-offs to be considered in
almost all aspects of condensed phase simulations.

...regards...dac

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu