Hi,
So quasiharmonic analysis (and the thermodynamic properties calculated
from it) really only works well if your system has one very
well-defined energy minimum (because of the underlying assumption of
harmonic oscillation). Mike Gilson et al. published a very nice
article a while back that goes into this in some detail:
https://pubs.acs.org/doi/abs/10.1021/ct0500904
Also, I've personally never found looking directly at the matrices to
be very enlightening. You may want to also try the 'projection'
command to projection your coordinates along the generated
eigenvectors, along with the various types of 'modes' analysis you can
do ('modes fluct' etc). See the cpptraj section of the Amber 17 manual
for more details and examples.
Hope this helps,
-Dan
On Tue, Feb 6, 2018 at 9:50 AM, Chetna Tyagi <cheta231.gmail.com> wrote:
> Dear Amber users,
>
> I have performed simulations on a small peptide (18 residues). I performed
> dihedral PCA on the results and analyzed correlation matrices to find out
> correlated motions amongst various sections of the peptide. I read the
> previous answers about the differences in PCA and quasi-harmonic analysis.
>
> Is it more beneficial to carry out quasi-harmonic analysis for such
> motions? I don't really understand the idea behind it but I performed the
> calculation. The matrix 'mwcvmat.dat' when plotted as a heatmap doesn't
> show anything interesting in comparison to dihedral PCA matrices.
>
> Also the file 'thermo.dat' with thermo chemistry calculation and the first
> 20 modes is produced. I am unable to apply the information to my simulation
> results. Does it show stability in simulation? What does the various energy
> terms calculated for each mode signify?
>
> I am attaching the heatmap of the matrix.
>
>
> [image: Inline image 1]
> Any help or advice is appreciated.
> --
> Best wishes
> Chetna
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
-------------------------
Daniel R. Roe
Laboratory of Computational Biology
National Institutes of Health, NHLBI
5635 Fishers Ln, Rm T900
Rockville MD, 20852
https://www.lobos.nih.gov/lcb
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Feb 09 2018 - 05:00:02 PST
