Calculating PMF from a constant velocity SMD simulation!

From: Gungor Ozer (
Date: Sat Apr 05 2008 - 19:10:06 CDT

Dear all,
I've done a series of constant velocity SMD simulations and now trying to
compute the average PMF of these trajectories. As far as I get from the two
Schulten papers ("Calculating potentials of mean force from steered
molecular dynamics simulations" in JCP & "Free energy calculation from
steered molecular dynamics simulations using Jarzynski's equality") and from
the manuals this is done;

1) by first calculating the work => integrate f*v*dt over time elapsed.
2) then by applying the Jarzynski's equality.

My questions:

1) Since any SMD simulation is not perfect, and not exactly following a
constant velocity path, shouldn't we use f*dx in eq.1 with dx being the
exact displacement of the steered particle?
2) f in eq 1 is a vector, how are we supposed to integrate a vector*scalar?

and a different question.

3) I am using TCL script to calculate force vector:

 set ff [expr $k*($nx*($r0x-$rx)+$ny*($r0y-$ry)+$nz*($r0z-$rz) +
 set fx [expr $ff*$nx]
 set fy [expr $ff*$ny]
 set fz [expr $ff*$nz]

ff being the force derived from the potential given U = 1/2k [ vt - (r-r0)*n
] where r is the current position, r0 is the initial position and n is the
direction vector.

Does my TCL script seem to calculate correct amount of work?

Any help to any of the questions will be much appreciated.


Gungor Ozer
Georgia Tech

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:49:21 CST