AMBER Archive (2005)

Subject: Re: AMBER: 1-4 scaling factors

From: Thomas E. Cheatham, III (cheatham_at_chpc.utah.edu)
Date: Fri May 20 2005 - 14:43:34 CDT

> Is there a reason why the 1-4 vdw and 1-4 electrostatic scaling factors,
> sncb and scee, are able to be set by the user? Under what physical

We have been debating whether or not this power should reside with the
user rather than have it set within the prmtop. The likely reason that
the user can set these values is probably legacy from force field
development/testing even before AMBER 4.1 (where it is more convenient to
simply change the input file rather than develop a new prmtop for each
scale factor change).

> One of the developers pointed out that the values are wrong if they are
> changed from the defaults because amber force fields were fit to those
> values so it wouldn't be amber if they were changed. So would the artificial

"AMBER" is a suite of programs for biomolecular simulation; it is not a
force field and in fact there are many force fields for "AMBER" or that
people may refer to as AMBER and there are also non-"AMBER" force fields
that can be run with the AMBER code. The Weiner et al. older force field
(parm91?) distributed with AMBER used SCEE=SCNB=2.0; Cornell et al.
(parm94) and more recent variants use SCEE=1.2/SCNB=2.0. The carbohydrate
force field (GLYCAM04?) uses no scaling factors for the 1,4's (SCEE=1.0,
SCNB = 1.0). Clearly these values have to be set, but you are correct
that the mdin file may not be the place since people can get burned. A
classic example from my past is when people applied distance dependent
dielectric models by accident with explicit solvent. There are a number
of ways you can mess up a simulation via the non-judicious choice of input
parameters (such as time step too large, shake tolerance too large,
pressure/temperature coupling too rapid, not removing center of mass
motion if necessary, forgeting to wrap PBC coordinates in long runs, etc,
etc).

However, as the developers have discussed, a good argument for NOT having
the 1-4 scaling factors specified by the user (and putting it in with the
force field in the prmtop/LEaP) is when you mix force fields such as
protein and carbohydrates; what 1-4 scaling factor is used in this case?
Currently there is no easy way in the input file to say, scale the protein
and not the carbohydrate.

For now, as always, it is recommended that users be careful and check that
the appropriate scale factors are set in the input file. Perhaps the
community will have this issue resolved for AMBER 9.

\ Thomas E. Cheatham, III (Assistant Professor) College of Pharmacy, Depts of
| Medicinal Chemistry and of Pharmaceutics and Pharmaceutical Chemistry
| Adjunct Asst Prof of Bioengineering; Center for High Performance Computing
| University of Utah, 30 South 2000 East, Skaggs 201, Salt Lake City, UT 84112
|
| tec3_at_utah.edu (801) 587-9652; FAX: (801) 585-9119
\ BPRP295A / INSCC 418 http://www.chpc.utah.edu/~cheatham

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu