On Tue, Sep 17, 2013, Ryan Muraglia wrote:
>
>
> One of my sanity checks was to run the TI scheme "backwards," meaning that
> in the groupfile, I flipped the order of the sander calls.
Are you sure that everything (initial coordinates, initial velocities, etc.)
is the same in the two runs? The usual debugging rules apply here: run
a short (say 20-step) MD run, with ntpr=1 and lambda=0.5. Do you see the
mirror-image behavior? Are the potential and kinetic energies the same
at the first step? At later steps? Does anything change if you set ntt=0?
Lots of things can cause two trajectories to get out of sync, but you can
probably track it down with a number of very short runs. TI in sander *is*
just a matter of running on a mixed potential surface and collecting
statistics, so things should match up.
...good luck...dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Sep 17 2013 - 13:30:02 PDT
