[RASMB] ellipsoids of revolution

[Borries Demeler]

> 
> Hi all
> 
> I have a very basic question. How does one get from determining the a/b ratio to
> the gross molecular dimensions of the protein molecule. The molecular weight
> should place a limit but I am a bit fuzzy as to how exactly this is carried
> out. 
> 
Mitra,
keep in mind that ellipsoids (for which a/b ratios can be specified)
are just one of an infinite number of possible interpretations of a given
f/f0 ratio. There is NOTHING in your sedimentation experiment that says a
particular ellipsoid (say, prolate ellipsoid) is preferable over another
(say, oblate ellipsoid). Both would give different a/b ratios for a given
f/f0 and volume, but are equally (un-)likely interpretations of the real
thing. Here is an example: a spherical molecule could have an unordered
loop that extends from the globular domain. As such it might act much like
a parachute, giving the protein a high f/f0 value. You could interpret
this with a long rod model, but clearly you would miss the overall
dimensions of the protein by much, given that it is mostly globular.

So these ellipsoids are just possible shapes. To get the real shape you
need to do higher resolution analysis like NMR, cryo-EM or X-ray. 
If you want to compare possible shapes for a given MW-f/f0 combination,
you can use the simulation modules in UltraScan. They will give you
the dimensions of a sphere, prolate ellipsoid, oblate ellipsoid and long
rod models.

-Borries