Eigenvalue, Frequency, Fluctuation Plot

[Shayan Monabbati]

I am comparing nma from bio3d to Nomad-Ref, and it appears that bio3d only creates 1 plot for fluctuations while Nomad-Ref creates a fluctuation plot for each mode. So I am wondering what the fluctuation plot in "plot(modes, sse = pdb)" is representing.

Also, I may have a misconception that high eigenvalues corresponding with low frequency modes? Both the eigenvalue and frequency plots have an increasing trend, so if you could clarify the reason for the similar trends that would be extremely helpful.

[Lars Skjærven]

Hi Shayan, The default behavior of function fluct.nma() is to calculate the average over all modes weighted by the corresponding eigenvalues (which is what you'll find in the modes object after running nma).

You can also specify individual modes, or a subset of modes:

m = nma(read.pdb("hel"))

# fluctuations based on all modes (default)
fa = fluct.nma(m)

# mode 7 (first non-trivial mode)
f7 = fluct_7 = fluct.nma(m, mode.inds=7)

# first 20 non-trivial modes
f720 = fluct.nma(m, mode.inds=7:26)

Note that the first non-trivial mode has index 7 (there are 6 trivial modes).

Hope it helps. Lars

[Shayan Monabbati]

Hi Lars,

Just wondering what is the units on the scale of the y-axis? Nomad-Ref shows in cRMS Angstroms and the trend is the same using fluct.nma() but the scale seems to be much smaller (0-4 on Nomad-Ref) vs (0 to 0.8 on bio3d)

[Lars Skjærven]

Å**2. Note that determining the amplitudes from NMA is highly unreliable. There are no amplitudes associated to each mode - only directionality.

[Shayan Monabbati]

So high eigenvalues do not correspond with low frequency modes?

[Lars Skjærven]

It should be the opposite yes.