From: John Philo <jphilo@mailway.com>
  To  : rasmb@bbri.harvard.edu
  Date: Fri, 30 Apr 1999 08:00:46 -0700

RE: Molecular weight (Mw)

The earlier responses from Walt, Bo, and Jeff did not address a couple
of issues raised by your query, so I'll add my two cents worth.

First, none of the others have actually directly addressed the issue
of the accuracy of the D values that are obtained. Comparisons to
simulated data (with or without realistic noise) show that my SVEDBERG
program and/or the finite element fitters by Bo or Peter Shuck will
give D values accurate to 1% or better. By the same criterion fitting
Gaussians to g(s*) gives accuracies of 2%-10%.

In practice with real proteins that are truly a homogeneous single
species (i.e. pharmaceutical grade material) I have found that
SVEDBERG gives D values that are consistent with the known sequence
mass within 2-5%, and the same should hold true for the finite element
fitters. The real data tend to over-estimate D and under-estimate M.
The reasons for this are not 100% clear, but it is my belief that part
of the reason is failure to account for the influence of the layer
near the walls and windows where flow is inhibited.

Having said that, I would disagree with the reviewer's statement that
you should calculate D "to test the accuracy of the reported s and the
Mw from sedimentation equilibrium". I wouldn't view the velocity
fitting to get D as a test of the accuracy of Mw---there is no doubt
that equilibrium is more accurate. I think what the reviewer really
means is that if the mass implied by the velocity fitting agrees
within expected errors with the equilibrium results (and the sequence
mass if it's not a glycoprotein) then that is strong evidence for
treating both the velocity and equilibrium data as a single species.

John Philo


-----Original Message-----
From: Houphouet Hy Yarabe [hyarabe@unix1.sncc.lsu.edu]">mailto:hyarabe@unix1.sncc.lsu.edu]
Sent: Thursday, April 29, 1999 9:21 PM
To: rasmb@bbri.harvard.edu
Subject: Molecular weight (Mw)


I reported Mw by sedimentation equilibrium and S values determined by
sedimentation velocity as part of a paper I've submitted to Analytical
Chemistry. The paper got some good reviews.  However, one of the
reviewers
wants me to consider adding values of diffusion coefficient (D) which
can
be obtained from the width of sedimenting boundaries measured in the
sedimentation velocity.  With D values, he adds, Mw results can be
calculated using the Svedberg equation and compared with Mw obtained
via
sedimentation equilibrium.  This comparison, he concludes, would
provide
an excellent test of the accuracy of reported s and Mw (sedimentation
equilibrium) values.

I aggree with him on the principle.  However, in my mind, the
diffusion
coefficients calculated from sedimentation velocity lack accuracy.
Could
somebody give me some input to address this reviewer.

"Struggle to know more ....."

Yarabe

Louisiana State University
Baton Rouge, LA 70803
Tel: 225-388-3919
Fax:  225-388-3971