AMBER Archive (2000)

Subject: gibbs ile->val energies

From: Steve Edmondson (steve_at_maxine.biochem.siu.edu)
Date: Wed Oct 25 2000 - 09:26:45 CDT

I am using gibbs (AMBER 5) to do FEP/TI of a ILE to VAL in an ACE/NME
blocked
pentapepride (no charge) in a water box. The input file from interface
is
shown below.
gibbs
temp / target=300.0 / initial=0.0 / constant=temp / coupling=split / \
     tau_solute=0.20 /limit_velocity = 10.0 /maxlimit=2
pbc / pressure / all_solute_nb / target=1.0 / tau=0.4
timestep = 0.10000E-02
pairlist / update=25 / cutoff=8.0
force_field /elec14_scale=1.2
dielectric = 1.0
shake / all / tolerance = 1.0D-6 / retry = 5
intragroup / non_bond / 1_4 / internals
mixing / vanishr = 0.05
entropy/d_temp=0.5
constraints /energy /ti_method=pf
free_energy_components /group /breakdown /steps=99999
group /residues = (-1,7) /title = "all solute res."
group /residues = (8,9999)/title = "water"
steps = (neq+ncoll) /runs=(nwindow+1)
report /steps=(neq+ncoll)
windows / start_lambda=1.0 / width = (1.0/20) / equil= neq /coll= ncoll
/ \
     decrease /ti

The forward free energies reported in the output are about 18. Based on
calculations I have seen in the lterature, I was expecting a much
smaller dG.
The largest contributions to dG are from intragroup electrostatics of
the
perturbed ILE (determined by omitting the intragroup option). But the
perturbed partial charges undergo only modest changes. What am I missing

here?

Stephen Edmondson
Medical Biochemistry
Southern Illinois University