[gmx-developers] re: Energy and Foreign Lambdas and Hamiltonian Replica

[Michael Shirts]

Hi,all-

> I'm fairly concerned about exchanging based on linear extrapolation of
> dV/dlambda. In my experience, dV/dlambda can be pretty nonlinear,
> especially in the case when soft core potentials are used -- even if
> lambda spacings are pretty close. I'm thinking the risks of doing
> things this way may outweigh the benefits, and that it would probably
> be better to hold off until foreign lambda energies are in that branch
> also.

I'm going to have to second this.  For a linear interaction potential,
this is of course 100% correct.  For a non-linear potential, this will
not end up being right.  I'm not sure it even would obey detailed
balance.  So we should wait until the foreign lambda code is in.

I have quite a bit of experience testing such code (since I've put it
in a GROMACS 3.1.4 version already), and would be happy to test
functionality out -- once it actually compiles :)  So keep me updated,
Erik!

Cheers,
Michael Shirts
Research Fellow
Department of Chemistry
Columbia University