<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=iso-8859-1">
<META content="MSHTML 6.00.2900.2523" name=GENERATOR></HEAD>
<BODY style="MARGIN-TOP: 2px; FONT: 8pt Tahoma; MARGIN-LEFT: 2px">
<DIV><FONT size=2>rasmb</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>I seriously doubt that you will find a much better way to
calibrate temperature than the CoCl2 approach Walter published. Trying to
use solubility introduces factors like concentration & nucleation effects
into the mix. It also only allows a measurement at one
temperature.</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>At the intro & service Beckman courses in Palo Alto at the
end of one of the days they half-fill two sectors of a prep rotor with vacuum or
drive oil, spin overnight, & put a thermometer into the solutions to check
temperature. Its one point per day but I guess a good way to check even
for prep centrifuges.</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>I certainly think this standardization is a good idea, & I
am glad to participate, although I must stress again I calibrate temperature and
we measure density, and density*vbar is a much larger source of error. So
get a density meter while you're at it!</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2> I would also encourage you to recognize this is not the
experts coming to the rescue. Any physical measurement requires
calibration and careful regulation of temperature. This is a fact of life,
not a special event that will somehow resolve the problem for everybody &
all machines. You must wait for temperature before hitting start, & if
you want to do careful work you must calibrate. (I asked for a poll
earlier - if you have calibrated temperature on your XLA/I send me a
note!)</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>The suggestions to do replicates is not something unique to
this event. You must always do replicates, as a function of concentration,
to understand the behavior of your system. The sample Arthur sends to us
will presumably be nonassociating, low charge, homogeneous, & in a buffer
near the isoelectric point. For typical samples association,
heterogeneity or nonideal behavior are the sources of deviation. In
my lab multiple users typically achieve 0.1 - 0.3% reproducible S values on well
behaved noninteracting systems, & this applies to joint analysis with
sedfit, DCDT+ and SEDANAL. (For monomers the analysis method should not be
that critical - but it is the holidays, so get in the spirit & sell
away.) I also honestly think users doing their own analysis should be part
of the experiment, for no reason other than setting a standard & expecting
the field to meet that standard. I think I have now crossed the PC line -
</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>happy holidays</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV> </DIV>
<DIV><FONT
size=2>-------------------------------------------------------------------<BR> Dr.
John J. "Jack" Correia<BR> Department of Biochemistry<BR> University
of Mississippi Medical Center<BR> 2500 North State Street<BR> Jackson,
MS 39216<BR> (601)
984-1522
<BR> fax (601)
984-1501
<BR> email address: <A
href="mailto:jcorreia@biochem.umsmed.edu">jcorreia@biochem.umsmed.edu</A>
<BR> homepage location: <A
href="http://biochemistry.umc.edu/correia.html">http://biochemistry.umc.edu/correia.html</A><BR> dept
homepage location: <A
href="http://biochemistry.umc.edu/">http://biochemistry.umc.edu/</A><BR>-------------------------------------------------------------------<BR> <BR> <BR></FONT></DIV></BODY></HTML>