Re: [AMBER] regarding remd

From: Daniel Sindhikara <>
Date: Mon, 26 Jul 2010 23:24:14 +0900

  Geometrical distribution of temperatures is optimum assuming constant
molar heat capacity.
This distribution may be non-ideal for systems with sharp phase transitions,
but it's still
a good guess to start. Optimizing the temperature distribution is
non-trivial, but a good way to check
is that you have* reasonable acceptance ratio.
Here are some references regarding the T distributions (not a complete

Kofke, D. A., On the acceptance probability of replica-exchange Monte Carlo
trials. *J. Chem. Phys. **2002,* *117* (15), 6911-6914.

Trebst, S.; Troyer, M.; Hansmann, U. H. E., Optimized parallel tempering
simulations of proteins. *J. Chem. Phys. **2006,* *124* (17), 174903.

Zheng, W. H.; Andrec, M.; Gallicchio, E.; Levy, R. M., Simulating replica
exchange simulations of protein folding with a kinetic network model. *Proc.
Natl. Acad. Sci. U. S. A. **2007,* *104* (39), 15340-15345.

Nymeyer, H., How efficient is replica exchange molecular dynamics? An
analytic approach. *J. Chem. Theory. Comput. **2008,* *4* (4), 626-636.

Also, *an even acceptance ratio does not mean detailed balance*. Detailed
balance is only achieved
after convergence (and if the proper Monte Carlo criterion is used).

So my advice is to start with a guess, and tweak by hand according to the
acceptance ratios if necessary. If you have
a sharp phase transition, you may want to check out the Trebst et al paper
(and I believe there are subsequent follow-ups).


On Mon, Jul 26, 2010 at 10:29 PM, Jorgen Simonsen <>wrote:

> Hi,
> I went through the tutorial for REMD and the article "Replica-exchange
> molecular dynamics method for protein folding". After reading a couple of
> article on the subject I am more confused about how to chose the
> temperature
> distribution. In general it should be chosen geometrically and with the
> degrees of freedom in mind - for the tutorial the following temperatures
> are
> chosen:
> 269.5, 300, 334, 371.8, 413.9, 460.7, 512.9,570.9
> as I would like to try it on a couple of different peptides some of them
> also have 272 degrees of freedom where I presume that the temperatures used
> in the tutorial will be the same for those but others have more degrees of
> freedom hence I would need to change the temperature to the corresponding
> dof so could anyone guide me to where the calculations for the temperatures
> are being performed?
> I ran one of the simulation for 100 ns in implicit solvent and get
> acceptance between the different intervals ranging form 0.17 to 0.19 which
> seem that the detailed balance is fulfilled which I interpreted that the
> temperature range for the 272 degrees of freedom is acceptable.
> _______________________________________________
> AMBER mailing list

Dr. Daniel J. Sindhikara
Institute for Molecular Science
AMBER mailing list
Received on Mon Jul 26 2010 - 07:30:05 PDT
Custom Search