AMBER Archive (2005)

Subject: Re: AMBER: Rep Exch. suggestion

From: Guanglei Cui (cuigl_at_csb.sunysb.edu)
Date: Mon Apr 18 2005 - 08:37:37 CDT

Hi, Brent

I'm sorry for the poor and lazy design of the REM code (I'm assuming you
are using the multisander REM). When I ran REM, I used to triple check
the mdins and made sure the only difference was the temp0. Including the
sanity check in sander requires certain MPI calls because of the way
mdins are read in. This may be included in future releases (more
significant changes on REM implementation may be coming as well from
other AMBER developers). On the other hand, writing a simple REM error
checking script might solve the problem more easily and quickly. Once
again, sorry for your suffering ...

Regards,

Guanglei

Brent Krueger wrote:
> AMBER Gurus,
>
> Just had a devil of a time trying to sort out a problem with a Replica
> Exchange run. Turns out that there was a typo in one of the input files
> such that two replicas were running at the same temperature. So,
> everything ran great until the first exchange was supposed to happen
> then the job hung. More importantly -- no useful error message or output.
>
> Perhaps R.E. runs could do a little sanity checking of the input prior
> to running?
>
>
> Cheers,
> Brent
>
> __________________________________________________________________
> Brent P. Krueger phone: 616 395 7629
> Assistant Professor fax: 616 395 7118
> Hope College
> Department of Chemistry
> Holland, MI 49423
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu