Re: AMBER: CHARMM vs. AMBER ff literature

From: Thomas Cheatham <>
Date: Wed, 11 Apr 2007 15:21:14 -0600 (Mountain Daylight Time)

> Is there a good paper or link someone can point me to that describes the major
> differences between CHARMM and AMBER Force Fields as they apply to nucleic
> acids?


A brief discussion is in a paper by Cheatham & Young, Biopol. (2000/2001)
56, 232-256 (which I will send directly to Seth) and another by
Reddy/Leclerc/Karplus, Biophys J (2003) 84, 1421-1449.

In terms of the model, the same empirical potential (roughly) is being
used, however the parameters were generated in a different manner (i.e.
different charge model, different vdw, different torsions, different
geometries, etc...).

In term of performance and in general (for simulations in explicit
solvent), for DNA on short timescales (< 10 ns) both the CHARMM all27 and
parm9X variants all perform moderately well. Issues with all27 are less
groove width dependence than expected and very rapid base pair opening; no
population of gamma=trans and possibly some other minor sub-states of the
backbone. With parm9X, there is an issue with anomalous sub-states of
(a,g)=g+,t; this leads to progressive deterioration of the structure in
longer runs (5+ ns). A fix is the parmbsc0 force field announced
previously on this list from the Orozco group.

For RNA, everything is peachy if the simulations are short; if the
simulations are continued for a "long" time (where long depends on the
force field and ranges from 2-200 ns), characteristic loop structure or
domain-domain interaction is lost with both force fields.

There are many reviews out there that highlight the performance of each
force field independently; for reviews on "AMBER", see the different ones
by Cheatham, Beveridge, Orozco, Laughton, Sponer, ... and for reviews on
"CHARMM" see different ones by Nilsson, Norberg, MacKerell, Feig, ...
Most of these paint a positive picture, and most of these are relatively
"old" in that they talk about simulations on the 1-10 ns time scale which
are not long enough to clearly expose deficiencies.

I would advise that more work is needed to improve *all* the available
nucleic acid force fields and recommend that if you are trying to model
unusual structures with modified force fields, you need to carefully
validate your results...

\-/ Thomas E. Cheatham, III (Assistant Professor) College of Pharmacy
-/- Departments of Med. Chem. and of Pharmaceutics and Pharm. Chem.
/-\ Adjunct Asst Prof of Bioeng.; Center for High Performance Computing
\-/ University of Utah, 30 S 2000 E, SH 201, Salt Lake City, UT 84112
/-\ (801) 587-9652; FAX: (801) 585-9119
\-/ BPRP295A

The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Sun Apr 15 2007 - 06:07:11 PDT
Custom Search