Re: [AMBER] Parmchk2 and dihedral parameters

From: Andreas Gavrielides <>
Date: Tue, 6 Dec 2016 13:02:49 +0000

First of all thank you very much for your very valuable advice. In the second question to be more specific, I mean is it ok to use the general dihedral for all the parameters involving the -c3-c3- atoms as atoms 2 and 3 and not take into account the specific dihedrals used or is it adviced to use the specific dihedrals produced by parmchk2?

Thank you again in advance

Andreas Gavrielides

Doctorant au CIRIMAT-INP Toulouse

SURF Equipe


 4 All??e Emile Monso BP 44362,

 31432 Toulouse Cedex 4

From: David A Case <>
Sent: Tuesday, December 6, 2016 1:45 PM
To: AMBER Mailing List
Subject: Re: [AMBER] Parmchk2 and dihedral parameters

On Tue, Dec 06, 2016, Andreas Gavrielides wrote:
> I am using AmberTools 16 to assign charges to my molecules and to fit missing parameters for the GAFF using parmchk2. Parmchk gives the following output:
> 1) Why are in some cases two parameters given for the same dihedral?

Many torsion potentials are expressed using more than one term: each term has
a different periodicity (a different value of "n" in the cos(n*phi) part.)

> 2) Can the general dihedrals (e.g X-c3-c3-X) be used in all dihedrals
> following the type or should I follow the parmchk suggestions?

I don't understand your question here...can you be more specific?

> 3) How can I recalculate parameters for dihedrals using the atomic
> charges calculated with Antechamber?

At the antechamber/gaff level of approximation, torsional potentials are
independent of the charge model. There are several tools in Amber that would
let you explore this is more detail, such as paramfit and mdgx. These involve
carrying out quantum chemistry calculations, then fitting the results to
molecular mechanics potentials. In a simpler exploratory fashion, you could
just switch between sets of charges (using the same torsion parameters) and
plot energy vs dihedral angle: this will give you an idea of how sensitive the
torsion potentials are to the charge model. Some types of torsions will be
much more sensitive than others.


AMBER mailing list
AMBER mailing list
Received on Tue Dec 06 2016 - 05:30:02 PST
Custom Search