Re: [AMBER] the problem of charge

From: FyD <>
Date: Sun, 29 Apr 2012 09:40:51 +0200

Dear Cao Yang,

> I am a user of AMBER in China. My system is a protein complex with
> phospholipid membranes. The particial charge of DPPC is calculated

MEP are computed using a QM program and charges are usually derived
from the MEP during an additional fitting step (using for instance the
RESP program). Indeed, even if the QM program provides with the MEP a
set of charge values (ESP) they are generally not suitable for MD
simulations as they are not equivalenced (chemical equivalencing for
chemical equivalent atoms lead to identical charge values for atoms
that are chemically equivalent: this is achieved during the fitting
stage(s) by using specific charge constraints).

> The charge of my ligand is calculated in AMBER with AM1.
> I load the pdb file into tleap and find the whole charge of my
> system is +5.6. So I add 6 Cl- which change the charge to -0.4
> rather than zero. I wonder why the total charge is not zero. So My
> question is:
> Whether the charge equaling -0.4 be resonable in MD simulation? If
> not, how should I make the system neutral?
> Thanks for your time!

- The total charge of a whole molecule such as DPPC should be an
integer value.
- If you generated a molecular fragment from the DPPC molecule one
could imagine a non-integer charge value for a molecular fragment, but
in this case the sum of the total charges of the different molecular
fragments constructed should lead to an integer charge value.

You can find two examples of lipids in R.E.DD.B. This corresponds to
the work of S. Abel. See:

In F-92, dodecylphosphocholine (DPC) is constructed from 3 elementary
building blocks... This is quite simple to adapt this approach to the
DPPC case.

You might also look for in the literature for already available parameters...

regards, Francois

AMBER mailing list
Received on Sun Apr 29 2012 - 01:00:03 PDT
Custom Search