Re: [AMBER] QMMM NEB reaction pathway calculation with AMBER

From: Jason Swails <>
Date: Sun, 29 Jun 2014 20:52:49 -0400

On Sat, Jun 28, 2014 at 8:50 PM, Teerapong Pirojsirikul <>

> Dear AMBER users,
> I am wondering whether QMMM NEB calculation in AMBER can be applied to the
> study of a reaction pathway associated with bond breaking/forming. As I
> understand, for this kind of reaction, we would end up with 2 completely
> different molecules (end points) and this means there would be 2 different
> topology files containing structural data for each of them. However, as
> documented in AMBER tutorial, it seems that 2 end points are required to
> share the same topology file. Does this mean we cannot use NEB calculation
> with the bond forming/breaking type of reaction?

‚ÄčA couple comments here.

1) When you are running a QM/MM calculation, all classical MM terms (bonds,
angles, torsions, charges, and vdW parameters) between two atoms in the QM
region are ignored. Therefore, as long as you never try and simulate the
product state with the reactant state topology file using only a MM
potential, there should be no problems using the same prmtop for all NEB

2) I believe the only hard requirement for NEB is that all beads have
exactly the same number of atoms in exactly the same order in every
topology file.

In any case, (1) is the most important here. If you think about simulating
a "simple" chemical reaction using QM/MM without using NEB (so you only
have one 'bead', perhaps using umbrella sampling), the same topology file
is used to simulate the whole reaction pathway. And that works precisely
_because_ all MM terms are ignored between QM atoms.


Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Sun Jun 29 2014 - 18:00:02 PDT
Custom Search