Re: [AMBER] Generation of velocities depends on conformation?

From: Thomas Cheatham <>
Date: Wed, 7 Nov 2018 05:15:05 -0700 (MST)

> I plotted the temperature after the first step of MD and observed something
> that puzzles me: the initial temperature is configuration dependent. By
> that I mean that for a given conformation, all 200 temperatures are around
> 400K. For another, the temperatures are around 350K. For others, around
> 300K. After 1ps of simulation, all temperatures are randomly distributed
> around 300K.
> Is it a known behaviour? I am also surprised that while asking for a
> temperature of 300K, the initial temperature is at 400K. Can someone please
> explain me how the velocities are initially assigned to each atom?

The velocities are assigned independent of the forces based on a
Maxwell-Boltzmann distributed targeting a particular temperature. What the
actual velocities will be after the first MD step depends on the forces
and the randomly assigned direction of the velocity.

Imagine a bond at equilibrium (no force) compared a highly strained
(compressed or extended) bond. If at equilibrium, net velocity should be
300K (if assigned at 300K). If stretched or compressed, depending on
initial velocity assignment, initial temperature could be higher or lower
than 300K.

If you want to avoid the large temperature spread, minimize the structures
or do low temperature MD first... A question to ponder is does it matter
if the temperatire is low or high at beginning, i.e. does it influence the
results? Having high force structures at higher than expected initial
temperature points out that the structures are relatively more strained.


AMBER mailing list
Received on Wed Nov 07 2018 - 04:30:02 PST
Custom Search