Hello Lammps users,
Is there anyway to do FEP with LAMMPS in a convenient way, or is there even a pre-made command for calculating and printing energies for two charge-states while running on a linear combination of them defined by a lambda? I cannot find much in the archives, but think it would be of general interest.
Yes, running trajectories, collecting snapshots, and afterwards calculating energies is doable, but it requires a lot of scripting - which I would like to avoid if it is available elsewhere - and it seems less practical.
Best wishes,
Mats.
Hello Lammps users,
Is there anyway to do FEP with LAMMPS in a convenient way, or is there even
a pre-made command for calculating and printing energies for two
charge-states while running on a linear combination of them defined by a
lambda? I cannot find much in the archives, but think it would be of general
interest.
the basic infrastructure exists (running multiple replica or "partitions"),
but i don't know of any direct implementation. there are multiple ways
how you could go about how to implement it.
Yes, running trajectories, collecting snapshots, and afterwards calculating
energies is doable, but it requires a lot of scripting - which I would like
to avoid if it is available elsewhere - and it seems less practical.
nod. it would definitely more useful, if you'd implement
it as a part of the "replica" package.
cheers,
axel.
You can also look at the fix adapt command which has the
ability to define the lambda parameter for some potentials.
People have used it to do free energy calculations.
But as Axel said, if there are more automated methods/comands
we could introduce for common modes of doing FE, we
could work with people to add them.
Steve