AMBER Archive (2005)

Subject: Re: AMBER: Glycam-parameters and antechamber

From: kkirschn_at_hamilton.edu
Date: Tue Sep 20 2005 - 15:37:16 CDT


Hi Mike,
     Your approach sounds fine for generating a rough estimation of
the results. I would expect that the partial atomic charges on the
sugar residue to be different than what is used in Glycam04 (ie. RESP
calculation with a weighting factor of 0.010 done at the HF/6-31G*//
HF/6-31G* level of theory). Realize that the partial atomic charges
determined in the manner are not consistent with the methodology used
in the development of the Glycam04 force field parameters. These
charges are inconsistent in two ways, the level of theory used and
the fact that they are not an average from an ensemble of conformations.

     An alternative low cost approach would be to determine the
partial atomic charges of BZ (capped with a methyl group) using HF/
6-31G*//HF/6-31G* RESP 0.010. This is a general "fast and dirty"
method for any special modification. Assuming you have made a prep
file for the modified carbohydrate using antechamber I would then do
the following steps:

1) Replace all of the BZ partial atomic charges in the antechamber
generated prep file, excluding the methoxy capping group (not just
the methyl group), for the HF/6-31G*//HF/6-31G* RESP 0.010 values.

2) Replace all of the carbohydrate partial atomic charges in the
antechamber generated prep file for the charges found in the Glycam
prep file that was distributed with AMBER for your carbohydrate (ie.
alpha galactose if you are adding BZ to alpha galactose) [also can be
obtain from the Glycam website]. Pay attention to the partial atomic
charge on the two linkage oxygens - you will have to use the
appropriate prep file (eg. 24la.prep if you are linking BZ at the 2
and 4 position of alpha galactose). These two oxygens should have a
"ester oxygen type" partial atomic charge. (ie. ~ -0.4 versus a
hydroxy oxygen type of ~ 0.7)

3) Modify the linking carbon on BZ (ie. the one attaching itself to
the carbohydrate oxygens) such that the overall charge on the new
prep file is 0.194. Thus, when you terminate your disaccharide with
a methoxy group you will have an overall formal charge of zero. I
would expect this to be a relatively small adjustment since the
partial atomic charge for the carbon was determined using BZ capped
with a methyl group (ie. the carbon is in a similar electronic
environment).

     You will then have a decent set of partial atomic charges,
lacking only in the fact that they were not determined using an
ensemble of conformations. Keep in mind that Glycam04 does not
contain any parameters for aromatic systems. You will have to load an
additional force field (ie. gaff) to obtain these parameters, and
make sure the atom types on the BZ are labeled correctly for this
force field. Thus, there will be some errors in the simulations due
to the fact that these are two independently generated force fields.
If you use a protecting group that is aliphatic (ie. tert-butyl
group) then you can use the Glycam04 force field by itself.

Hope this helps.
Cheers,
Karl
____________________________________
Karl N. Kirschner, Ph.D.
Visiting Assistant Professor of Chemistry
Hamilton College, Clinton NY 13323
____________________________________

> Hello dear amber community,
> just a brief question.
> I would like to perform an MD simulation of a disaccharide (GlcNAc-
> Gal) that carries some modifications (basically protecting groups
> like Bz) at two positions of the Gal moiety. Obviously I would like
> to still make use of the GLYCAM parameters...
> I know that in principle I should calculate RESP charges for this
> new moiety and so on...but I don't have so much time & PC-power...
> I was therefore thinking to use antechamber, calculate AM1-BCC
> charges, editing the atomtype manually (and change them to the
> glycam-conventions) and then, after having setup the system,
> perform my MD simulation to obtain some clue about the conformation
> of this modified disaccharide.
> Would that be at least a rough estimation of the result I would by
> doing it "correctly"??
> Thanks for you help
> mike
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu