Re: [AMBER] Questions about restarting REMD simulations

From: Daniel Roe <>
Date: Fri, 10 Oct 2014 07:48:16 -0600


There are a few possibilities here. One is that you are not actually
restarting your runs (i.e. irest = 0) and the temperatures in two or
more of your input files are the same. The second is that you are
restarting (irest=1) but your restart files do not have temperatures,
so they are using the temperatures in your input files and the first
case is in effect. The last is that you are restarting and two of your
restart files do actually have the same temperature. This can happen
e.g. if the previous run crashed in the middle of writing restart
files and some were left at the previous restart point. You can check
the time stamps on your restart files to see if this is the case. This
is usually more of an issue if you are writing restarts very

One way you can check the temperatures in your restart files (if they
are ascii) is:

 head -2 restart.* | sort -g -k 3 | tail -n X

Replace 'restart.*' with any filename mask that will select all of
your restart files and 'X' by the number of restarts you have.

If you have duplicate temperatures in your restarts you will just have
to start a new run, which may not be too bad.


On Fri, Oct 10, 2014 at 12:25 AM, 王漠野 <> wrote:
> Hello,
> I met a problem in my REMD process,could anyone help me please?
> I am trying to restart a replica-exchange MD simulation. But this is the message I get in the files.
> In other replays someone said that check the header in the restart files to see if they are all the same time stamp.I read my restart files I found that there was nothing wrong with them.What should I read in the files so that I can find the errors? And my job was killed during the process.With this condition,what should I do with my restart files and how to resart the REMD process normaly? I am using the Amber10 now.
> Thank you!
> Wangmoye
> University of Chinese Academy of Sciences
> --
> 王漠野
> _______________________________________________
> AMBER mailing list

Daniel R. Roe, PhD
Department of Medicinal Chemistry
University of Utah
30 South 2000 East, Room 307
Salt Lake City, UT 84112-5820
(801) 587-9652
(801) 585-6208 (Fax)
AMBER mailing list
Received on Fri Oct 10 2014 - 07:00:02 PDT
Custom Search