Re: [AMBER] pre-processing of trajectories

From: Hannes Loeffler <>
Date: Thu, 24 Jun 2010 11:04:25 +0100

On Thu, 24 Jun 2010 12:50:46 +0530 (IST) wrote:

> Thanks a lot for the reply.
> When I view my trajectories in VMD, the dimeric protein moves to
> different position within the solvent box but does not come out of
> the box or the system does not 'break-up'. When I check for
> equillibrium using standard parameters mentioned in the tutorial such
> as total enegy, density, temp, etc and also rmsd and radius of
> gyration, an equilibrium seems to have been reached. However, when I
> do PCA using ptraj and plot the projections on the top components, it
> fluctuates rapidly as shown in the figure attached, one reason for
> which can be the periodic imaging artefact. When I do center and
> image the plots attached improves and shows equillibration with the
> projections being distributed around a mean value. But if there is no
> imaging artefact, why should this happen at all?

For the purpose of the PCA what you have to make sure is that your dimer
stays intact at all times. If that is true for all time steps of your
simulation re-imaging wouldn't have any effect because you RMS fit the
structure from every frame to the reference structure. Of course, you
must make sure that your reference is an intact dimer too.

To check what is happening you could just write out the re-imaged
trajectory and compare it visually to the original one.

AMBER mailing list
Received on Thu Jun 24 2010 - 03:30:03 PDT
Custom Search