From: JOHN PHILO <JOHN.PHILO@amgen.com>
  To  : rasmb <rasmb@bbri.eri.harvard.edu>
  Date: 23 Jun 1995 09:09:56 U

Re: w^2*t

        Reply to:   RE>w^2*t

Michael Jacobsen asked:

>5. What of other analysis packages currently available-do these also >rely
on the integral obtained by the XLA?

My SVEDBERG program uses the w^2*t values stored in the data files.  As Bo
Demeler has already pointed out, this is the correct way to derive the
position of the boundary, taking into account finite movement during
acceleration and/or low speed scans.  Thus the sedimentation coefficients
derived by SVEDBERG should be correct even if you pause at low speeds for a
long time.  On the other hand, SVEDBERG also derives the diffusion
coefficient from the boundary spreading.  Since diffusion is related more to
time than w^2*t, the boundary shapes and derived diffusion coefficients will
be altered if you make significant changes in rotor speed during the run.  
This is not a problem if you just pause a few minutes at 3000 rpm to take a
preliminary scan (as I always do) and then accelerate to your final speed. 
It would be a 
problem, for example, if you ran at 30K for a while to follow a fast-moving
component and then accelerated to 60K to follow a slow-moving component.


>6. Does the "time" format stipulated by the user (viz. HOLD, TIME, >w2t)
bear any affect upon what value of the integral of w2t "comes up" >above each
scan?

This setting only determines how the XL-A decides when (or if) to stop the
run.  It does not effect the data files.

I hope this helps clarify things.

John Philo, Protein Chemistry
Amgen
jphilo@amgen.com