On Sun, Feb 8, 2015 at 7:05 PM, Emmanuel <emmanuel.salawu.gmail.com> wrote:
> Dear Amber Users,
>
> Can you please enlightening me on how I can output (i.e. dump to file) the
> forces acting of each atom during MD simulation using Amber14. I found in
> the mailing list (http://archive.ambermd.org/201308/0097.html) that this
> feature would be (and, hopefully, is now) available in Amber14.
>
Yes it is available. You can write a separate NetCDF or ASCII 'force'
trajectory *or* you can write a NetCDF trajectory with coordinates and/or
velocities and/or forces. Have a look at the "ntwf" variable (page 287 in
the Amber 14 manual). If you set it to -1 (as I recommend), then the
forces are saved alongside the coordinates in the trajectory file (and the
file is simply 2x bigger). If you set it to another (positive) number, you
can specify the name of the force trajectory file via "-frc <mdfrc>" (mdfrc
is the default name).
One thing to note here, though, is that the forces and coordinates will be
one frame off from one another -- the way sander/pmemd work (loosely) is to
compute forces from coordinates, use those forces (with the velocities) to
propagate the coordinates, correct the velocities, then write out the
coordinate/force/velocity trajectories. Ergo, the coordinates that are
printed come from the conformation *after* the ones for which the forces
were computed. So if you go back and compute the forces for each frame in
a trajectory, they will not match the forces printed at each time step
exactly. Just a heads-up.
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Feb 08 2015 - 18:30:03 PST
