Hi JJ,
> 1) Should ED analysis be performed only on the segment of trajectory wherein
> the protein's RMSD has equilibrated, am I right? Because I have the notion
> that harmonic analysis of a trajectory can only be performed when the
> protein is undergoing fluctuations about a minimum. In other words once rmsd
> stabilizes, then one can perform a harmonic analysis.
This depends on what you're after. First of all, eigenvector analysis
makes no assumptions on the shape of the (multivariate) distribution.
However, subsequent analysis may require harmonicity. Unfortunately,
the leveling of the RMSD has nothing to do with harmonicity of
motions, and less with harmonicity of projections on eigenvectors.
You'll have to check the distributions for the eigenvectors you're
interested in.
> 2) Which would be a better marker to see if the protein has equilibrated to
> the simulation environment the RMSD of the protein or the projection of
> motion along Eigen vectors?
The RMSD is a univariate projection of the complex state of a protein.
That means you loose a lot of information. Moreover, since the RMSD is
only a distance measure, it is very crude in the sense that you can
have a lot of different configurations for a certain distance,
especially when the distance gets larger. When looking at the
distributions of the projections of the trajectory on a number of
eigenvectors, you'll regard more information and more specific
information.
> 3) When I plot the root mean square fluctuation (r.m.s.f.) of the amino
> acids am I plotting the motions (of c-alphas) orthogonal to the principal
> component, Eigen vector 1? and would I wrong in saying that these regions of
> the protein which have large fluctuations, (say greater than 2 angstroms),
> do experience anharmonic fluctuation?
No and yes. When you plot the rmsf you'll plot that part of the rmsf
that is explained by the n-th eigenvector, not by the orthogonal ones.
And again, thihs has nothing to do with anharmonicity, it could also
be harmonic motion.
> If so how can I pick up anharmonic fluctuations?
Look at the distributions.
Maybe it's also good to do more background reading on principal
component analysis. I've given a few elaborate answers on this list in
the past, which may be worthwhile and there's quite a number of
documents available online that explain the technique from a more
statistical point of view, that may aid in overseeing the
possibilities and limitations of PCA.
By the way, 7 ns is very short for this type of analysis. You're
unlikely to have convergence with less than 15 - 20 ns, although it
depends on the size of your system.
Hope it helps,
Tsjerk
--
Tsjerk A. Wassenaar, Ph.D.
Junior UD (post-doc)
Biomolecular NMR, Bijvoet Center
Utrecht University
Padualaan 8
3584 CH Utrecht
The Netherlands
P: +31-30-2539931
F: +31-30-2537623
___
gmx-users mailing listgmx-users@gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please search the archive at http://www.gromacs.org/search before posting!
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-requ...@gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php
