Re: contraint failure in RATTLE algorithm

From: Jim Phillips (jim_at_ks.uiuc.edu)
Date: Fri Nov 07 2003 - 10:33:01 CST

Hi,

I see these warnings at the start of the run:

Info: Configuration file is waku.conf
Warning: Following are the variables which were set in the
Warning: configuration file but which are NOT VALID
Warning: run
Warning: reinitvels
Warning: minimize
Warning: set

In NAMD 2.5 these would be fatal, but earlier versions muddle on anyway.
Clearly this is not a good thing, which is why it is fixed now.

These errors indicate that NAMD was built without Tcl, and is interpreting
these Tcl interface commands as config file options. Therefore NAMD is
going straight to dynamics without doing any minimization at all.

Since you appear to be using a straight Linux-i686 version, you may be
able to download the precompiled NAMD 2.5 binaries and just drop them in
yourself. These binaries are well optimized for both AMD and Intel CPUs,
and you should see a nice speed boost from going to 2.5 as well.

-Jim

On Thu, 6 Nov 2003, ELENA JEAN LEVIN wrote:

> Hi Brian,
>
> Thanks for the suggestions. Upgrading to version
> 2.5 is unfortunately not an immediate option, as
> our sys admin person is busy for the next few
> days. We tried gradually increasing the
> temperature, as you suggested, using the following
> script:
>
> reassignFreq 1000
> reassignTemp 25
> reassignIncr 25
> reassignHold 300
>
> numsteps 50000
> minimize 1000
> reinitvels 0
>
> However, we got the same error message. We tried
> running the entire simulation at 1 Kelvin and
> still had the RATTLE error, so I don't think that
> a sudden jump in temperature is the problem. Do
> you have any other ideas?
>
> Thanks,
> Elena
>
> ----- Original Message -----
> From: Brian Bennion <brian_at_youkai.llnl.gov>
> Date: Thursday, November 6, 2003 2:10 pm
> Subject: Re: namd-l: contraint failure in RATTLE
> algorithm
>
> > Hello Elena
> >
> > First suggestion, try the final 2.5 version of
> NAMD, if nothing
> > else it
> > will clean up the log file.
> >
> > Second, and I hope Jim or somone else will
> correct me if I am
> > wrong, but
> > the protein-rattle algorithm can't handle a
> single energy pulse of
> > 313K.
> > I have received similar warnings when first
> using namd without
> > first
> > trying to
> > heat the protein/water system gradually.
> >
> > Brian
>
>
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:37:06 CST