AMBER Archive (2006)

Subject: Re: AMBER: Why "ho" and "hw" VDW's radius are set to 0 in "gaff.dat" ?

From: David A. Case (
Date: Thu Jul 06 2006 - 20:48:12 CDT

On Thu, Jul 06, 2006, Christophe Guilbert wrote:

> When minimizing one of my ligand which include one "ho" atom type, I
> found that it collapsed (superimposed) with its nearest atom generating
> a crazy electrostatic energy.

It is in general necessary to use SHAKE (ntc=2) when you have hydrogens
with zero vdW radius. Basically, this keeps the OH distance fixed, and the
vdW of the O atom prevents the H from getting too close to a C. This won't
work if the carbon is bonded to the oxygen (i.e. you have a C-O-H group that
is causing the problem), but in most cases, it should help.

You could, of course, complain to Junmei Wang, who followed Peter Kollman's
lead in this regard; Peter, in turn was following ideas put forth by Bill
Jorgensen. But Bill was thinking of rigid molecules, and in any event has
already received plenty of complaints on this score. Peter can't defend
himself anymore. But maybe the current Amber force field developers will
think about fixing this problem as they improve/modify the force fields.
Certainly, gaff is evolving, and changes here may happen.


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