Tom,
We developed and parameterized am1-bcc to have charges equivalenced on
topologically equivalent atoms. We have not yet examined the behaviour of
the am1-bcc implementation in antechamber; of course we use our own code
from when we developed the method. We take charges from a single calculation
and simply average those of topologically equivalent atoms; this can be done
either before or after the bcc corrections have been applied to the am1
charges.
Averaging over multiple conformations is not something currently part of
"best practices" but it is a good way to deal with the small charge
variations with conformation seen with am1. Note that these variations are
"real" from the standpoint of the am1 hamiltonian (i.e. reflecting genuine
changes in electronic structure), not dominated by numerical instability as
in the case of ESP-fit charges. Thankfully they are usually small unless
you're breaking conjugation or something.
Here are some "best practice" issues:
1. The am1 charges used should come from an am1-optimized structure; valence
bonds and angles especially need to be optimized.
2. Intramolecular H-bonds in the am1-opt structure MUST BE AVOIDED unless
you deliberately want the significantly-perturbed am1 atomic charges that
result. Recommendation: get your am1 charges from an am1 opt on an extended
conformation. Use an internal Z-matrix (xyz translate via Babel) and freeze
torsions to non-H atoms if you must in order to avoid internal H-bonds.
3. For accurate work (approx or exact free energy calcs), use only one
consistent set of charges amongst conformers, especially with molecules
carrying a net formal charge.
4. Average charges on topologically equivalent atoms, e.g. methyls,
methylenes, {monosubst. and para-disubst} phenyls, {carboxylate, sulfone,
sulfonamide, phosphate, and phosphonate} oxygens.
Those are pretty much it from what we have seen so far.
I hope this helps,
Christopher Bayly
-----Original Message-----
From: Tom Bishop [mailto:bishop_at_tulane.edu]
Sent: Monday, September 23, 2002 5:29 PM
To: amber
Subject: antechamber best practices/charge equivalence w/ AM1-BCC
Dear amber,
I am using antechamber to assign types and charges to a set of small
organic compounds
and would like some information about proper usage.
We wish to generate charges with am1-bcc and have followed examples in
manual.
Doing so we notice that charge equivalences are not enforced.
Examples are given to demonstrate charge equivalencing for a two stage
fitting of a gaussian generated ESP.
Is charge equivalencing w/ AM1-BCC desirable?
Does one use multiple conformations to achieve this?
What are best practices? Any pointers to examples would be greatly
appreciated.
Thanks in advance.
TOm
--
-----------------------------------------
Thomas C. Bishop, Ph.D.
Joint Faculty Appointments Program Professor
Center for Bioenvironmental Research
-----------------------------------------
Tulane University Office
-----------------------------------------
Dept. of Environmental Health Sciences(SL-29)
School of Public Health & Trop. Medicine
1430 Tulane Avenue
New Orleans, LA 70112
Phone: 504-988-6203 Fax: 504-585-6428
bishop_at_tulane.edu
-----------------------------------------
Xavier University Office
-----------------------------------------
Division of Basic Pharmaceutical Sciences
College of Pharmacy
7325 Palmetto Street
New Orleans, LA 70125
Phone: 504-485-5106 Fax: 504-485-7954
tbishop_at_xula.edu
-----------------------------------------
------------------------------------------------------------------------------
Notice: This e-mail message, together with any attachments, contains information of Merck & Co., Inc. (Whitehouse Station, New Jersey, USA) that may be confidential, proprietary copyrighted and/or legally privileged, and is intended solely for the use of the individual or entity named on this message. If you are not the intended recipient, and have received this message in error, please immediately return this by e-mail and then delete it.
==============================================================================
Received on Tue Sep 24 2002 - 07:49:56 PDT
