AMBER Archive (2006)

Subject: Re: AMBER: SHAKE - ntc = 3

From: David A. Case (case_at_scripps.edu)
Date: Fri Jan 06 2006 - 19:26:55 CST

On Fri, Jan 06, 2006, Jennie Thomas wrote:
> I am running a simulation of NO3- in water using the following frcmod file:
>
> This is the additional/replacement parameter set for nitrate ion NO3-
> MASS
> ON 16.000 1.000
> NO 14.000 0.000
>
> BOND
> ON-ON 500. 2.197
> ON-NO 500. 1.269
>
> ANGLE
> ON-NO-ON 100. 120.0
> ON-ON-ON 0. 60.
> NO-ON-ON 0. 30.
>
> DIHE
> ON-ON-ON-ON 2 0.0 0.0 2.
> ON-ON-ON-NO 2 0.0 0.0 2.
> ON-ON-NO-ON 2 0.0 0.0 2.
> ON-NO-ON-ON 2 0.0 0.0 2.
>
>
> NONB
> NO 1.88 0.1700
> ON 1.80 0.160
>
> In my input file, I have specified ntc=3, so all of the bond lengths
> should be fixed. However the O-O bond length (averaged over time) is
> about 0.1 Angstroms shorter than the length specified in the force
> field

Do you actually have bonds defined between all pairs of oxygen atoms?
(Use rdparm to list all bonds in the molecule to make sure).

You should almost always set ntf = ntc, but I don't think(?) this should be
the cause of your problem. Still, setting ntf=3 will remove all bonds
from the energy term, and should help in debugging.

If you really have defined bonds between all pairs of atoms, this is a very
difficult challenge for SHAKE, but it should work. And, you don't need to
run 100 ps to see what is happening: try a simulation of 20 steps or so,
saving the trajectory on every step, and look closely at the structure on
every step.

...good luck...dac

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