Re: Accumulating free energy change for applying colvar restraint

From: Jérôme Hénin (
Date: Mon Oct 07 2013 - 09:06:15 CDT

Hi Randy,

> My questions:
> 1) This lambda schedule is specified individually for every collective
> variable - how can that be? What if the schedules or number of dynamics
> steps are not consistent with each other?

To be very precise, the lambda schedule is associated with one restraint (the harmonic {...} block). Each restraint may act on one or more colvars. If several independent restraints are defined, then the program does not restrict them to have the same lambda-schedules, it is up to the user to come up with a sensible restraining scheme.

> 2) In the sample namd input file (rest-01.namd) the only action that deviates
> from a straight dynamics simulation is to turn on the colvars -
> colvars on
> colvarsConfig
> - there is no other specification connected to FEP. In fact, in running
> this, I see no 'special' output connected to the variables or the changing
> hamiltonian.

I don't know the specifics of the tutorial, but usually this "free energy of restraint" calculation is performed entirely at one end-point of the FEP (lambda = either 0 or 1). In that case, a regular (non-alchemical) topology describing either end-point can be used, and there is no need to invoke the FEP code.

> "The blunt, pathetic reality today is that a little old lady has died, who in
> the winter of her life had to water roses alone under police supervision. If
> you behave like there's no such thing as society, in the end there isn't. "
> -- Russell Brand on Margaret Thatcher

Very nice quote.


This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:21:45 CST