[gmx-users] Free energy calculation

[Justin A. Lemkul]

Fabian Casteblanco wrote:
> Hello all,
> 
> I am setting up a free energy calculation (drug from full coulomb+vdW
> in solution --> drug with only vdW in solution --> dummy drug in
> solution).
> 
> After reading most of the papers, I understand that you need
> significant overlap from the energies for each intermediate point to
> overlap so its best to have many intermediate points from Lambda=0 to
> Lambda=1.  The drug molecule I have is a bit complex but I wasn't sure
> if using too small of an intermediate could have a bad effect on the
> free_energy calculation.  I know Justin Lemkul said in its tutorial
> that Lambda=+0.05 should be good for most but I decided to go with
> Lambda=+0.02.  Could this have any negative effect other than taking a
> longer time?  Also, how does one come up with the best soft-core
> potential parameters to use?  Is it ok to use the one from the Methane
> in water tutorial?
> 

The settings from the methane tutorial should basically be defaults, which 
generally work well.  Unless you have specific reason to change them, they 
should be fine for most purposes.

I don't know if one can have "too much" overlap, but using BAR even 0.05 is 
overkill, I've found.  For methods like TI you would likely need 0.05, or even 
less in some cases, but BAR converges better.

-Justin

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================