08-26-96

The original TIP3P model (J. Chem. Phys., 1983, 79, 926-935)
is a rigid three-site one, where the hydrogens are simple
point charges (no Lennard-Jones parameters).  Though this
is a rigid model, bond and valence terms have been added for
use in MD simulations (J. Phys. Chem., 1987, 91, 3349-3358).


The OPLS/TIP3P force field utilizes a Coulomb + 6-12 LJ
format where the combining rules are expressed as geometric
means. NBFIX statements are used to yield accurate water-water
interactions.  At this time, nothing has been done for solute-
water interactions.

Because of the problem with splitting dipoles, this model
was parametrized with and requires the use of GROUP-based
cutoffs.  Also, the use of a SHIFTing function for the cutoff
of electrostatic interactions is very bad for fluid simulations;
switch to truncation.

If you're using one of the following parameter sets:

topallh22x.lip/parallh22x.lip
topallh22x.nuc/parallh22x.nuc
topallh22x.pro/parallh22x.pro
toph11.wat/param11.wat
toph19.rcv/param19.rcv
toph19.sol/param19.sol
param6a.inp

then comment out any lines dealing with the TIP3P or
TIPS3P model (it is *not* the original TIP3P model).
These lines will have 
       OT
       HT
       ot
       ht
as atom types.  Then, use the tip3p.topology and tip3p.parameter
files here.

If you're using 

opls.topology/opls.parameter

everything is OK.