Re: [AMBER] How to choose ntc value?

From: Jason Swails <>
Date: Sat, 14 Sep 2013 23:53:55 -0400

On Sat, Sep 14, 2013 at 1:55 PM, X Sun <> wrote:

> Dear all,
> I am doing a MD simulation of interaction between ionic liquid and
> cellulose. With assist of tutorial B1, the system is easy to run normally.
> But I am confused about the value of ntc, in tutorial B1, the author said
> ntc=2 would constrain the movement of hydrogen atom, and that would not
> affect the result in simulation B1. But I focus on the interaction between
> two compounds and the comformation change and hydrogen bonds formed among
> them. I think the movement of hydrogen atom is very important for me. How
> can I find a rule or instruction to choose a proper ntc and ntf?

I think you misunderstand what this flag is doing. This flag specifically
constrains the hydrogen-heavy atom bond to be a specific length -- it does
not prevent the hydrogen atom from 'moving'.

The bonds between hydrogen atoms and heavy atoms are typically quite stiff
(i.e., they have a high force constant), and given the very light reduced
mass of the bonded pair yields vibrations with a very high frequency. The
time step you can choose is dictated by the highest frequency motion in
your system to avoid integration error. Roughly speaking, these
hydrogen-heavy atom covalent bonds have a frequency at least double that of
all other bonds in the system, so you can use a time step that is twice as
large if you constrain those degrees of freedom.

The reason this is such a commonly employed technique is that those motions
with the fastest frequency are often the least interesting motions in the
system (it is the slow, large-scale conformational changes that are
typically the most interesting).

You will get a better explanation (with less hand-waving and more, better
examples) by skimming an introductory computational chemistry book.


Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Sat Sep 14 2013 - 21:00:03 PDT
Custom Search