[RASMB] questions

[Daryl Bosco]

Hello, I have just started performing analytical ultracentrifugation 
experiments and have some basic questions for more experienced users.

I have collected an initial data set on a Beckman XL-A and analyzed data 
using Origin 4.1 16 bit software with the Self-Association Model.  I am 
looking at a monomer/ dimer equilibrium between two DIFFERENT proteins, 
where the Kd has been estimated by other methods to be 15 uM.  At higher 
concentrations (above 15 uM) I see there is a larger dimer content.  Taking 
3 data sets at 5, 15 and 60 uM, I get an apparent Ka of 1.9 in absorbance 
units (Origin software)

My questions:

To convert the Ka to a Kd in concentration units, do I take the reciprocal 
of the Ka (1/1.9) and divide that by the extinction coefficient and path 
length?  ie   0.53/(1.2cm)(19000cm-1)or is the calculation more 
complicated?  This will give ~ 20uM in agreement with the other methods.

I am going to use the NONLIN software for self-association in order to 
(again) calculate the Kd for protein A and protein B.  Is the self 
association model a good model for monomer/ dimer equilibrium for different 
proteins (as opposed to protein A forming a dimer with 
itself---homodimerization)?

How many data sets does one acquire to calculate an accurate kd?  Is it 
similar to a binding curve where you want to start below the kd and have 
points WELL above the kD (SATURATED)?  I do not expect 60 uM to be fully 
saturated, but I do get a reasonable kD from 3 points (listed above) and if 
I go any higher the Absorbance within my cell will be too high.

Finally, how do you calculate speed and concentration factors for the 
NONLIN software (in the dialog box before the catual fitting)?

I appreciate any feedback you could give-

thank you

Daryl Angela Bosco