Re: [AMBER] iwrap on very long simulation

From: Jason Swails <>
Date: Mon, 08 Sep 2014 09:07:11 -0400

On Mon, 2014-09-08 at 18:02 +0530, wrote:
> Dear Prof. Case,
> You have specified ntxo=2, but in the Amber12 manual ntxo= 0 (Unformatted
> (no longer recommended or allowed: please use formatted restart
> files)
> OR
> =1 (default) Formatted (ASCII) are there.
> I didn't find an option ntxo=2.
> Is there such option ??

Yes. See the errata for the manual at

> I have also a question on iwrap. If we set ioutfm=1 then we don't have to
> set iwrap=1 even for long simulation. Am I write ?

If you set ioutfm=1 and ntxo=2, then you will never see the problem that
coordinates overflow fixed-size fields and replaces the coordinates with
***s. It's possible to run long enough that the single-precision
trajectory files may lose precision for the particles that have migrated
very far away from the original box, though (although you probably have
to run for a very long time for this to become a real problem).


Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Mon Sep 08 2014 - 06:30:02 PDT
Custom Search