Hi Pawan,

I can add two things. First of all, the score is simply the projection
of a point (conformation) onto the eigenvector. Second, extreme
scores/projections, be it positive or negative, are usually most
unlikely, thus corresponding to highest energy. For the  lowest
energies, you'd be looking for projections that occur most, thus
correspond to the modus of the distribution (hence 'modal').
I hope with this, you'll have enough to hold on to work your way
through. If it still eludes you, I'm sorry, but I don't think I can
help you much further to sort it out. Probably the best you can do is
find some texts from statistics about principal component analysis
(plenty on the web), and take some simple (non-MD) exercises. If
you've got the view on PCA straight, then turn back to MD. Oh, maybe
the paragraph on PCA in my thesis (see
http://dissertations.ub.rug.nl/FILES/faculties/science/2006/t.a.wassenaar/02_c2.pdf)
may also be a useful read.

Good luck!

Cheers,

Tsjerk

On Sun, Dec 12, 2010 at 3:40 PM, pawan raghav <pwnr...@gmail.com> wrote:
> Dear tsjerk,
> I am very sorry to say that your suggestions does not help but increase
> confusions only. I am highly obliged if you give me clear statements about
> the scores to understand this concept in depth. I have searched research
> article about this concept but failed to find even an single article.
> Neither Gromacs tutorial give any description about these scores, nor your
> PCA tutorial. I am really interested to understand this concept so help by
> giving me answers of my questions.
>
> eigenvector           Minimum           Maximum
>
>                  value       time      value       time
>
>       1      -9.162661       99.0   2.682097     9450.0
>
>       2      -2.728093     4695.0   3.272116      558.0
>
> I got clearcut idea about this line after read your tutorial "These are the
> maximum and minimum projections on the eigenvectors."
> What is the meaning of this line here? "They are very unlikely to correspond
> to energy minima, as minima will be modal."
> If the values in second column not the energy values they are scores only.
> Are scores directly corresponds to the energy minima for second column and
> energy maxima for 4th column on which they are based? I have seen that in
> second column there are negative values (Minimum) and in 4th column there
> are positive values (Maximum). In above example along the first eigenvector,
> Are -9.162661 value at time 99 ps corresponding to the structure at low
> energy during the projection and the value 2.682097 score corresponds to the
> maximum energy. If I am wrong then tell me, what are the criteria on which
> there scores are based?
> I am totally agree with your this concept and understand completely in this
> line "Think of a pendulum, projecting the position on the floor."
> But this statement confused me "The extreme projections actually correspond
> to states of higher (potential) energy". Which extreme projection
> corresponds to state of higher (potential) energy either minimum or maximum
> or both of them?
> Please reply me line by line so that I can understand this concept.
> --
> Pawan Kumar Raghav
> Bioinformatician
>
> --
> gmx-users mailing list    gmx-us...@gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
>



-- 
Tsjerk A. Wassenaar, Ph.D.

post-doctoral researcher
Molecular Dynamics Group
* Groningen Institute for Biomolecular Research and Biotechnology
* Zernike Institute for Advanced Materials
University of Groningen
The Netherlands
--
gmx-users mailing list    gmx-users@gromacs.org
http://lists.gromacs.org/mailman/listinfo/gmx-users
Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists

Reply via email to