[ccp4bb]: Summary: DLS: discrete dimer vs random assemlby

[diane h peapus <dp72@cornell.edu>]

***  For details on how to be removed from this list visit the  ***
***    CCP4 home page http://www.dl.ac.uk/CCP/CCP4/main.html    ***


Thanks to everyone who answered my DLS question!

There _does_ appear to be a ProteinSolutions message 
board.  It doesn't seem to be very active.

http://www.protein-solutions.com/cgi-bin/msgboard/login.html


Original question:
> I'm running dynamic light scattering (DynaPro99) and am
> wondering how to interpret what I'm looking at.  
......
> I'm wondering if anyone has any rule of thumb about how
> sharp a peak needs to be to call the solution homogenious?



Concensus answers:

-Look at the errors:  If they are large, redo the experiment.
baseline should be 1+/-0.003
count rate should be steady
SOS error should be less that 5

Make sure the routine to exclude bad data points is enabled!!!

(At the pull down menu) Tools -> Settings -> Data filtering

Dan Snyder of Protein Solutions recommended these limits...
Over SOS Error:  not needed, coupled to rest
Under baseline limit:	0.98
Over baseline limit:	1.02
Under Amplitude limit:	0.01
Ignor 1st # coeff:	4
Truncate at channel #:	120

To make these as default, set them with _no_ data set open.
If you set new limits with a data set open, the limits will 
apply to _only_ the open data set.  All previously taken
data sets need will contain the old limits... unless you
reset the limits and run.... 
(At the pull down menu) Analysis -> Recalculate All   
for each data set.


-Look at the polydispersion index: 
This is the percentage obtained by dividing the dispersity 
of your peak (how broad it is) by the hydrodynamic radius.

If it is less than 0.1 your solution is monodisperse.  
Some people went as high as 0.15 to say it is monodisperse.
Above 30% is a polydisperse solution


-Look at the Bi-modal distribution
If the polydispersion index is larger than 0.1, the bi-modal
distribution can sometimes tell you if there is a high MW 
aggregate that is contributing to the scattering.  The % of 
each component is listed.


Notes on applying info data to crystallization:

S. M. Roe reminds us that...
The peak will always come to a higher MW than the true 
MW, unless your protein is a perfect sphere, due to 
unaccounted for additional rotation friction. 

Giovanna Scapin found that...
A protein does have higher chances to crystallize if it is
monodisperse (which is not saying it will crystallize), 
but a low level polydispersity (2% or less) of aggregation 
in most cases did not make a difference.

Jeff Habel (with a paper soon in ACTA Crys D) says:
On several occasions I have crystallized solutions with a
Cp/Rh (polydispersion index) of 20-25%


Thanks again
diane


-- 
"You don't need to be smart to be president"
     --Republican Congressman J.C. Watts - at a February campaign 
       appearance on Bush's behalf. Washington Post, 6/11/00

"I think anybody who doesn't think I'm smart enough 
to handle the job is underestimating."
     --George W Bush - U.S. News & World Report, April 3, 2000