Re: [AMBER] Probelm with antechamber for resp charge derivation

From: FyD <>
Date: Wed, 23 Feb 2011 08:59:13 +0100

Rajesh Raju,

> I have attached the molecule in jpg format. The system is a salt whose
> charge is neutralized by 4 Br- anions. I have a couple of questions:
> [1] Do I need to include the Br- anions in making the paramter file..,
> but the exact position of the Br- is unknown...

I would first construct a model without Br-; Br- would be added by the
"addions" command of the LEaP program. MD will decide where the 4 Br-
will go and/or you could specify some input positions.

Once again, please, read in detail the paper of Cieplak et al. (JCC
1995) you will find many common points between the polymers described
and your project.
You can find also complementary information in the paper describing
the R.E.D. Tools (PCCP 2010).

> [2] when I optimize at HF/6-31G*, do i need to include the Br- anion..

I would first try optimizing the building blocks without Br-

> Also from literature, Br- aninons helps to maintain the stacked form of
> the two pyrene rings,
> But when I optimzed, without Br- aninons, the molecules moving away..

This is normal: you try to get a conformation observed in experimental
conditions, which are totally different to these used in geometry
optimization. You could try to modify the conditions used in geometry
optimization but you will never get exactly the same ones (in
particular if your molecule is docked into a protein).

Instead, you could identify the building blocks constituting your
molecule and you could use specific charge constraint during the
charge fit to reconstruct your initial/large molecule from these
building blocks. At the end, you use LEaP & its "sequence" command to
connect the different fragments generated.

- Advantages of this building block approach:
* You work on small elements and the corresponding geometry
optimization step is performed quickly
* The conformations of the building blocks are always fully defined -
and not based on the nose/foot of the user ;-)
* You do not favor any specific interaction within your initial/large
molecule because of the geometry optimization conditions used
* You can potentially generate an infinity of analogs for your
initial/large molecule; for instance, if you have 3 building blocks;
you can generate different sequences between these building blocks;
3-(2)2-1-(2)2-3 in your case; then 3-(2)n-1-(2)n-3 or with structural
adaptations *,# 3#-(2)n-1*-(2)n-3#
* You can potentially build any type of biopolymers once you
understood how the approach works

- Disadvantage of this approach
* The approach looks complex for new users; But I think you pay for
what you get! R.E.D. & now R.E.D. Server have been developed to help.
You can find examples in R.E.DD.B. & tutorials
* You need to determine the correct building blocks & the correct
charge constraints (INTRA-MCC and/or INTER-MCC keywords in a P2N file)
applied during the charge fit; The new Statistics module of R.E.D.
Server helps to identify bad approaches/building blocks/charge

Your particular case:
I think you could define 3 building blocks:
- charged_+1 imidazole with 2 connecting groups
- benzene ring with 2 connecting groups
- pyrene ring with 2 connecting groups

For benzene & pyrene, see:
For imidazole, see:

If you use R.E.D. III.4 (convenient to learn/start) you will be able
to use one/two INTER-MCC; if you use R.E.D. Server the number of
INTER-MCC is not limited.

When you use INTER-MCC or INTRA-MCC, chemical equivalencing in the
connecting groups has to be broken to limit the number of constraints
during the fit; this leads to lower RRMS values.

Do not use Ante_R.E.D. 1.4; it is limited compared to R.E.D.
Server/Ante_R.E.D. 2.0

I hope this helps.

regards, Francois

AMBER mailing list
Received on Wed Feb 23 2011 - 00:30:04 PST
Custom Search