Re: Accelerated MD

From: Jeff Wereszczynski (
Date: Wed Aug 29 2012 - 15:47:31 CDT

Hi Thomas,

How are you choosing your E and alpha variables? Here's an excerpt
from something I wrote before on this list:

In general choosing your parameters can take some trial and error. For the
amber force field we do have a rule of thumb that might be helpful. Run a
short classical MD simulation and calculate the average dihedral potential
energy (V_d) and the average total potential energy (V_tot). For a system
with Nres residues and Natom atoms, a good starting point for the dihedral
parameters may be:


And for the total you could try:

alpha_tot= N_atom/5

You should see an increase in the total potential energy of the
system, especially the non-bonded values. Could it be that you just
set your E_tot value too low? Perhaps you could post the aMD segment
of your input file, and a line from the log file listing your


On Wed, Aug 29, 2012 at 1:06 PM, Thomas Albers <> wrote:
> Hello!
> Does anyone on this list perhaps have experience with accelerated MD?
> We are looking at a short peptide in a membrane (30000 atoms total)
> and are trying to find suitable values for E and alpha in dual-boost
> MD.
> Here are the relevant average energies (kcal/mol) for classical MD:
> dihedral 4070.72, total -42381.53
> With accelerated MD we see:
> dihedral 6362.78, total -40156.92
> The dihedral energy is higher, as expected, because the system
> explores a greater amount of phase space, but it seems that the
> contribution to the increase in total energy comes almost exclusively
> from the dihedral energy. If I read this right there is almost no
> boost at all applied to the diffusive degrees of freedom. Am I
> understanding the data right?
> Thomas

Jeff Wereszczynski
Postdoctoral Scholar
University of California, San Diego

This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:22:27 CST