AMBER Archive (2008)

Subject: AMBER: gaff paramters

From: servaas michielssens (
Date: Fri Feb 22 2008 - 11:38:53 CST


I was making a new topology for a molecule that contains a P-N bound, I used the gaff.dat file to generate the new paramters. When I had a look at the new paramters the gaff parameters for the X-p5-n3-X in the gaff suprized me:
X -n3-p5-X 6 18.000 180.000 2.000

I expected:
1. the forceconstant value to be much lower
2. the phase to be 0
3. pn to be 3

So I decided to recalculated the parameter for the molecule in the attachement, I followed the strategie as in J.comput chem 24: 1016-1025, 2003.
Results point out that the function is nothing like the parameters in the gaff but a much more complicated function. It cannot be fit with 1 cosine function (which is no problem of course) and barriere is +/- 2kcal/mol (Vmax).

Their was also another problem:
o -p5-o 46.0 115.80 SOURCE3 17 4.6921 5.7902
O2-P -O2 140.0 119.90
Their must be something wrong with the force constant in the amber gaff.dat here (or my interpretation of them)...
same for:
o -p5-oh 43.8 115.03 SOURCE3 86 3.5330 4.9414


servaas michielssens

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