AMBER Archive (2006)

Subject: Re: AMBER: the energy and temperature both increasing during the simulation with NTT=0

From: David A. Case (case_at_scripps.edu)
Date: Tue Nov 21 2006 - 00:09:28 CST


On Mon, Nov 20, 2006, Xiaowei (David) Li wrote:

> I have observed that the total system energy and temperature were
> increasing in a 10ns production simulation (NVE, NTT=0) of a DNA
> decamer. The energy has increased from -63510 to -63410 kcal/mol while
> the temperature increased from 300 to 302 K.

Setting dt=0.002 is at the limit of energy stability for any sizeable system.
Try some (short?) runs with dt=0.001 if you like. You should see the RMS
fluctuations in total energy go way down (factor of four perhaps), and I think
that the energy drift will also be significantly better.

Also, setting nscm=1 seems unusual to me. It seems possible that this could
have some effect on long-time-scale energy conservation. It would certainly
be worth a try turning this off. Basically, you want to turn off anything
that interferes with Newton's equations.

...hope this helps....dac

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