[gmx-users] FEP and error estimation

[Tomas Kubar]

hi

we have implemented the "reverse cumulative averaging" technique for the
thermodynamic-integration simulations in gmx 3.3. this automated
procedure should provide both proper equilibration of the system at
every lambda value and controlled uncertainty of dG/dlambda (and thus
the error of resulting delta G) - Journal of Chemical Physics, Volume
120, Issue 6, pp. 2618-2628 (2004). maybe this could help you, maybe
even someone else? what do the gmx-users think?

tom

Maik Goette wrote:

> Thank you David
>
> I will discuss the problem with my colleagues first (and try to get an 
> appropriate solution), but maybe I will come back to your offer later. ;)
>
> Regards
>
> Maik Goette, Dipl. Biol.