[gmx-users] free energy and fatal error

[David Mobley]

Murat,

> I am using GROMACS 3.2.1.

I wrote the tutorial you are following. It is for GROMACS 3.3/3.3.1.
If I remember correctly, sc-power is an option that is new to GROMACS
3.3, hence the tutorial will not work in 3.2.1.  It is likely that the
"blowing up" you are experiencing is due to using soft core parameters
which are inappropriate.

> > I think, it has nothing to do with the error, you experience, but
> > delta_lambda should be 1/number of steps, if you want to do TI from 0 to 1.
> I follow the solvation FE tutorial by Dill group. These are the parameters that
> they provide

Maik was referring to the time-dependent scheme for changing lambda,
which I don't particularly recommend.

David