[RASMB] velocity sedimentation

[John Philo]

Pascal and RASMB,

Overall I think there is perhaps too much concern about the issue of scan
speed for absorbance scans with respect to DCDT analysis, which in the
extreme has led some people to believe this approach can't be used at all
(definitely false!). It is important to remember that the primary
determinant of signal/noise in the DCDT method is the magnitude of the
difference between the first and last scans used in the analysis, and that
is really determined by the masses of the species in the sample (the time
span must be limited to avoid peak broadening) and that difference is
independent of the scan speed. Once the time span from first to last scan is
fixed, the signal/noise only increases as the square root of the number of
scans, so even factors of 2 in scan speed have fairly small impact on the
overall signal/noise.

I do not recommend that you use replicates in velocity scans, and especially
not in the 'continuous' scan mode (as opposed to 'step' scan). In continuous
scan mode the slit is still moving as it takes the replicates, and thus the
readings from different radii are getting averaged together (a bad idea).
Further, in my view a spacing of .02 cm is far too large. For one thing,
that large spacing means the subtraction of time-invariant noise from
scratches or dirt on the windows (one of the primary advantages of the DCDT
approach) does not work well because the radial profiles are so sparse.

I'm not sure why you saw a significant difference between 2 versus 3
replicates; in theory this should make little difference, assuming you
increased the number of scans used in the analysis for the 2-replicate data,
since the scan speed was faster. (Remember to use the 'Test time range for
broadening' function in DCDT+ to help you decide how many scans to use.)

My recommendation is that you use the default settings for velocity runs:
continuous mode, .003 cm spacing, 1 replicate. To optimize things, you do
want to reset the default radial range (5.8 to 7.3 cm) so you scan from just
to the left of the meniscus down to ~7.10-7.14 cm, avoiding the region at
the cell base were solutes accumulate (i.e. don't waste time taking data you
can't use in the analysis). That will give a scan rate of about 80-90
seconds per cell.

Regarding rotor speed, the faster you spin the better the resolution among
different species, but you must balance this against loss of signal/noise
because you have fewer scans to analyze. Also for higher mass species at
some point high speed gives boundaries so sharp the diffusion coefficient
(and hence mass) cannot be determined accurately. In your case I would think
50-55 K is about right.

John Philo
  -----Original Message-----
  From: rasmb-admin at rasmb-email.bbri.org
[mailto:rasmb-admin at rasmb-email.bbri.org]On Behalf Of pascal egea
  Sent: Wednesday, April 10, 2002 5:53 PM
  To: rasmb at rasmb-email.bbri.org
  Subject: [RASMB] velocity sedimentation


  Greetings,

  I am currently doing some velocity sedimentation experiments using
absorbance optics on a 33 kDal protein and I am processing my data with
DC/DT plus.
  I have several questions to ask to the g(S) function specialist.
  first I was spinning at 50.000 rpm and scanning 3 replicates with a radial
increment of 0.02 this was resulting in a scan every 3:30 minutes (roughly).
  my g(S) function is very nice clearly indicating a single component system
with a S of about 2.3 S.
  As It seems that the more frequently you scan the better it is, I
therefore tried only 2 replicates and still a radial increment of 0.02 and
now it allows me to scan every 2 minutes (approximately). nevertheless my
function does not look as nice (this is the same sample). I also increased
the speed to 55.000 rpm.

  so is it a result of a lower number of replicates (more noise)?. Is it
best to scan more frequently with more noise versus less frequently with
more replicates?
  Is there for sedimentation velocity a really optimal speed (I know the
reference curve) or the faster you spin the better it is?

  Thanks in advance for your answers.

  Pascal Egea
  UCSF department of Biophysics and Biochemistry
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.rasmb.org/pipermail/rasmb-rasmb.org/attachments/20020411/00ff900f/attachment.htm>