AMBER Archive (2005)

Subject: RE: AMBER: Is the MD simulation normal

From: Yong Duan (duan_at_ucdavis.edu)
Date: Sun May 29 2005 - 11:46:44 CDT


Shulin,

You have an average Ca-RMSD below 1.0A! This is an indication of a
reasonable trajectory.

To understand the difference of the trend before and after 300ps, you
need to appreciate some basic condensed matter theory such as chaos and
random field. What it means behind these scary terms is that the events
tend to be related to time in a logorithmic scale.

You perhaps also should appreciate the term "probabilistic", meaning
things sometimes happen and sometimes do not, very much like gambling.
Things that happened may or may not because of something you did right
or wrong. It is just a matter of chance. Here, in simulation, you are
doing stat-mech on computer. You are doing probabilistic study of
biochemistry. That is why people do multiple simulations, just to make
sure whatever they observe is not because they are too lucky or too
unlucky.

At the ns time scale, it is indeed a good idea to do more simulations
and to run the simulations much longer. This way you can see a clear
trend. You need to make sure you change the initial random number seed
and do so just once at the beginning.

yong

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu