This is not an urgent question by any means - but I am curious....
Comparing parm91.dat and parm94.dat, the vdw epsilon parameter for S (and SH)
changes (slightly) from 0.20000 to 0.25000 with the note added in parm94.dat
(and later): "W. Cornell CH3SH and CH3SCH3 FEP's"
I have performed FEP calculations, but not with a goal of parameter development.
I would like to understand this a bit better.
I checked Cornell et. al. (J. Am. Chem. SOC., Vol. 117, No. 19, 1995) and the
"methods" section mentions:
"Free energy perturbation calculations for perturbing methanethiol
to methanol and dimethyl thioether to dimethyl ether were carried out
using the AMBER program and the slow growth method. Simulations
were run for 200 ps with a time step of 2 fs. SHAKEM was applied to
constrain all bonds and perturbed bonds were shrunk. Only the solution
perturbation was carried out (with TIP3P waterI8 and periodic boundary
conditions) and the intramolecular components were not included.
Calculations were carried out in both the forward and reverse directions.
The PMF correction was included to account for the free energy change"
but I don't see how to connect the use of FEP (with which I am somewhatfamiliar)
to the "0.25" (I could envision fitting .25 to a forward calculation from
lambda=0 to lambda=1... but in a "slow growth" procedure, I am not seeing how
this might be done)
(Searching for "thio" elsewhere in the 1995 paper only turns up discussion of
the bond stretch parameter.)
Does anyone know how Dr. Cornell got from .2 to .25?
Was it as straightforward as "trying" .25, and noting that the delta-delta G of
solvation energy was more in line with experimental values? Or, was there a
more elaborate fitting process?
(Apologies for such a longwinded email over .05 kcal/mol :) )
Thanks
Chris
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Fri May 06 2005 - 20:53:00 PDT
