Re: [AMBER] REMD: What can I do with much larger replicas than the number of processors

From: Qinghua Liao <>
Date: Wed, 7 Nov 2018 20:43:00 +0100

Hello Zhonghua,

For such a big system, T-REMD can be very computational demanding. I
would recommend

Hamiltonian replica exchange simulations (REST2), of course, it also
depends what sampling you want to enhance.

All the best,


On 11/7/18 4:22 PM, Zhonghua Xia wrote:
> Dear all,
> I want to do T-REMD on my protein (88 AA, 1438 atoms) with TIP3P water box (5938 water molecules), as well as a counterion Cl-. I learn from the tutorial that the number of replicas should be equal to the square root of the total number of atoms. In my case, about 120 replicas…
> Any help appreciated! My purpose is certainly to enhance sampling.
> Kind regards,
> Zhonghua
> _______________________________________________
> AMBER mailing list
AMBER mailing list
Received on Wed Nov 07 2018 - 12:00:02 PST
Custom Search