Re: AMBER: Nudget Elastic Band query

From: Francesco Pietra <>
Date: Thu, 5 Jul 2007 13:28:31 -0700 (PDT)

Yes, your point was clear to me, though my reply was quite incorrect. What I
should have asked, and I am asking now, is where to find a protocol - more or
less detailed - to reparameterize for my molecules gaff parameters. I have
command of a high level QM package and a parallel computer with 4GB/node that
can do that.

To be specific, the reparameterization via QM should involve the groups
concerned (bonds, angles, dihedral angles) as isolated units, or right in the
context of the specific molecule? It makes a big difference.perhaps to make the
approach impossible.

Also, do you mean reparameterize all parameters involved in my molecule or only
those "Calculated with empirical approach" through antechamber? Just to have a
general guideline, for the molecule I tried (98 atoms), antechamber showed
three types of parameterization:

MASS: nothing shown.

BOND: nothing shown.

ANGLE: only one "Calculated with empirical approach".

DIHE: two indicated "same as" another dihedral.

IMPROPER: for three "General improper torsional angle (2 general atom types)".

For eight: "Using default value"

Sometimes, like now, I wonder how mmff94 succeeds in giving about correct
answers for small conformational energies.

Thanks. Your advice about my question would be a precious guideline.

francesco pietra

--- Carlos Simmerling <> wrote:

> On 7/5/07, Francesco Pietra <> wrote:
> >
> >
> >
> > 90:10 population of conformers, ie 1.3 kcal/mol at rt, is an ideal
> > situation
> > for conformational studies (i.e NMR techniques can deal with so that one
> > is not
> > computing blind). Conformational energy differences are always in this
> > range
> > with natural products (or organic synthetic drugs). If MD can't deal with
> > such
> > energy differences, it will be of little use in this area and the
> > interaction
> > of with receptors. However, I vaguely remember of published MD dealing
> > with
> > such small energy differences. Perhaps I can find out the paper spending
> > some
> > time.
> true, but my point was just that automatic parameters may not be good
> enough.
> I think most people use antechamber when they need lots of quick parameters
> but for a single molecule where you want high accuracy you will probably
> want to
> at least check the parameters using high level QM.

____________________________________________________________________________________Ready for the edge of your seat?
Check out tonight's top picks on Yahoo! TV.
The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Sun Jul 08 2007 - 06:07:32 PDT
Custom Search