AMBER Archive (2004)

Subject: Re: AMBER: strange output of MD

From: Bill Ross (
Date: Wed Jan 07 2004 - 12:39:35 CST

> It really does matter with the restrt from minimization as input
> for MD.
> But I am still confused that I once used just the crd file instead
> of the restrt from minimization as input for MD and the output of
> the MD was correct. And the two molecule size are the same. Then
> why one is OK while the other fails? Would you please give some suggestion?


It depends on two things: how bad the starting coordinates are,
and the topology of the molecule. The initial VDWAALS=23565.9892,
and you chose random velocity assignment to give a starting
temperature of 300K, which caused problems. (Slow warming can
substitute for minimization, in that low-temp dynamics typically
minimizes better than minimization.)

The large initial vdw in the molecule points to overlapping
atoms; since the other potentials were reasonable this suggests a
topology that can fold into itself.

Possibly the random velocities were less lucky in the 2nd molecule;
in any case you could look at the energies of that md run for clues.


The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to