From: Jose Garcia de la Torre <jgt@fcu.um.es>
  To  : rasmb@alpha.bbri.org, djs17@york.ac.uk
  Date: mar, 28 mar 2000 13:16:44 +0200

Re: protein R_g and hydrodynamics

Dear Dr. Scott and RASMB colleagues:

The radius of gyration, R_g, and the hydrodynamic radius, R_h give not 
equivalent but complementary information on the size and shape 
of the protein. If the amount of effective hydration is the same for
scattering as for hydrodynamics, size effects are eliminated, and then 
any combination involving the R_g/R_h ratio is an indicator of shape only. 
One of such combinations is P_O = f / [sqrt(6) eta_0 R_g], where 
f = 6 pi eta_0 R_h is the translational friction coefficient of the 
macromomolecule in a solvent of viscosity eta_0, so that  
P_0 = sqrt(6) pi R_h/r_g. (see Eur. Biophys. J. 25, 361 1997) 
The upper limit is P_0 = 9.93 for spherical particles. Values of 
P_0 for non-spherical shapes can be predicted (i) for particles of 
arbitrary shape represented by bead models, using our HYDRO/SOLPRO 
computer programs (see our web page; address below) (ii) for ellipsoidal 
shapes, using procedures developed by Prof. S.E. Harding.

If you have a determined or estimated three-dimensional atomic-detail
structure of the protein, in the format of a PDB file, then there is
a nice possibility: you can use our new program HYDROPRO (Biophysical J.
78, 719, 2000; program available from our web page) to calculate values 
of R_h, R_g and other solution properties from the 3-D 
structure to be compared with the experimental ones. This approach is
particularly fruitful when you have two, as in your case, or more
properties available. 

With my best regards,    

Jose.