AMBER Archive (2007)

Subject: Re: AMBER: error due to exceeding left-hand side digital numbers in restart output

From: David A. Case (case_at_scripps.edu)
Date: Mon Jun 18 2007 - 13:06:16 CDT


On Mon, Jun 18, 2007, Cenk Andac wrote:
>
> I have been running a 300 ns MD for an RNA hairpin solvated by TIP3
> waters. somewhere around 50 ns of MD , I seem to get some errors with
> restart files so that the next sander routine does not start. After
> going through the coordinates in the previous restart file, I have
> noticed that some coordinates were filled with lots of stars, like
> ****.******* , instead of numbers.

There is not much you can really do here. Back up to the last good restart
file, and start again, setting iwrap=1 in the sander input. Going to netcdf
trajectory output would also eliminate this problem, but I think most people
(including me!) have been slow to make the change to the new format.

...regards...dac

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