[gmx-users] mean square displacement frequency

[Tsjerk Wassenaar]

Hi Greg,

So you want a distribution of MSD, right?

> I have trajectories which has been obtained from simulating 20,000 atoms in
> NVT simulation for 25ns.  I recorded the trajectory every 3ps which means
> that I have 60,000 data points per atom.

25000/3 = 8333

Where do these 60k per atom come from?

>I am trying to construct a
> histogram where y-axis is the frequency (# of atoms) and x-axis is the mean
> square displacement over 25ns.
>
> I think a simple way to achieve this by looking at individual atoms and
> calculate its msd over 25ns and put them in a bin. But there are 20,000
> atoms there and 60,000 data points per atom which makes me hard to do this
> on a spreadsheet -- even if I cut down the data points for each atoms by two
> orders of magnitude, I have to deal with 12 million data points.

g_rmsf gives you sqrt(MSD) per atom. That should bring you quite close
to what you need.
Also give the manual a read once (more); there's many more tools that
can do all kinds of things that may at some point be useful.

Cheers,

Tsjerk




-- 
Tsjerk A. Wassenaar, Ph.D.

post-doctoral researcher
Molecular Dynamics Group
Groningen Institute for Biomolecular Research and Biotechnology
University of Groningen
The Netherlands