On Mon, Feb 12, 2018 at 3:04 PM, Lee-Ping Wang <leeping.ucdavis.edu> wrote:
> I want to do this because the kinetic energies that are written to “mdout” and “mden” do not correspond exactly to the written trajectory frames (i.e. mdout/mden are shifted by one frame with respect to the trajectory and velocity-trajectory). I need the kinetic energies that correspond exactly to my position trajectory frames, and it seems the only way to do this is to recompute them from the velocity-trajectory (please correct me if I’m wrong).
This is only half right. Due the the leapfrog nature of the integrator
the velocities that get written are half a time step ahead of the
coordinates (i.e. coordinates are written at time t, velocities at
time t+0.5*dt), so in order to get the velocities at time t you also
need forces. I believe that one way to do this is:
Vn = Vh - 0.5 * dt * Fn / m
where V is velocity, n denotes time t, h denotes time n+0.5*dt, dt is
the time step, F is force, and m is mass. However, this also will not
be exactly what you get from sander because sander estimates Vn
differently and depending on what thermostat you are using there are
certain additional corrections that are used. Ultimately right now I
think the code in runmd.F90 (and corresponding code in pmemd) is the
only reference for how Amber estimates kinetic energy - we should
probably develop a formal FAQ for this at some point.
Other in the Amber community please correct me if I'm off base here. :-)
-Dan
--
-------------------------
Daniel R. Roe
Laboratory of Computational Biology
National Institutes of Health, NHLBI
5635 Fishers Ln, Rm T900
Rockville MD, 20852
https://www.lobos.nih.gov/lcb
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Feb 20 2018 - 13:00:02 PST
