[RASMB] systematic differences

Tom Laue Tom.Laue at unh.edu
Wed Sep 6 06:58:47 PDT 2006


Hi-
I thought I had broadcast my response to Jack, but hadn't. Here it is-
Since no sedimentation standards are used, every element (hardware, 
software, data manipulation) of AUC contributes to the final values 
obtained.
Hardware contributions that may contribute to cell-to-cell differences 
include the center position of the rotor holes, the alignment mark at 
each rotor hole, the centerpiece condition, which mold the centerpiece 
came from, the condition of the cell housing, window holders, window 
gaskets, etc. I'd hate to think the rotor holes are mis-machined. Of 
these possible contributions, my guess is that the which mold was used, 
the centerpiece condition and the cell alignment are the most likely 
cause of the differences.
We have not measured the sector accuracy of Beckman centerpieces, and I 
suspect that the values will vary with time as the molds are refurbished 
or replaced. Keep in mind, too, that any batch-to-batch variation in 
temperature at the curing step will affect the mold dimensions. The 
variation in width at the base of the mold tines can be checked, and 
variations of 125 microns can be detected. I do know that Spin 
Analytical will sell batches of centerpieces made from the same mold. Be 
aware, though, that using centerpieces from a single mold will increase 
the precision of measurements, but will not enhance accuracy. While we 
are on the topic of centerpiece accuracy, there is significant variation 
in the thickness of Beckman centerpieces, which directly affects the 
concentration measurements from the absorbance and interference systems. 
The dimensions of Spin Analytical centerpieces are checked individually 
and a QA document included for each one.
Best wishes,
Tom

Leech, AP wrote:
> Dear all,
>
> A quick back of the envelope calculation assuming the width of a sector
> is 3mm, makes Jack's 0.05 cm discrepancy equal to 18 mm^3 (18 uL) 
> difference
> in volume in the cell (for about 420 uL filling), which would correspond
> to a difference in width of ~0.125 mm along the solution column. Does
> anyone know if this is within manufacturing tolerance of the cells? It
> seems a little large. Perhaps the CAMIS people have an idea of the spread
> of cell measurements?
>
> I presume Jack tracked the discrepancy by not refilling the cells but
> shaking them up then moving them?
>
> Andrew
>
>
> John Philo wrote:
>> Jack, since I haven't seen any other responses to your message from last
>> week I will jump in and offer one theory.
>>
>> You haven't said how you are filling these cells so reproducibly that 
>> you
>> can measure this correlation between meniscus position and cell 
>> number. You
>> also haven't said what type of centerpieces you are using, but I will 
>> guess
>> they are charcoal epon. Then if that correlation is correct it means the
>> channel volumes are different for different centerpieces (the same 
>> volume
>> gives a different column height). That I think can only be true if the
>> center rib is bowed, probably due to a leak at run speed (one channel 
>> empty,
>> the other full). That is, I think this whole exercise is telling you 
>> that at
>> least two of your centerpieces are damaged.
>>
>> Also, regarding your "correction factor", how can you say which cell 
>> gives
>> the "right" sedimentation coefficient and which ones need correction?
>>
>> John
>>
>> -----Original Message-----
>> From: rasmb-bounces at rasmb.bbri.org 
>> [mailto:rasmb-bounces at rasmb.bbri.org] On
>> Behalf Of Jack Kornblatt
>> Sent: Thursday, August 31, 2006 12:28 PM
>> To: rasmb at server1.bbri.org
>> Subject: [RASMB] systematic differences
>>
>>
>> This note relates to systematic between cells in the AUC.
>>
>> The problem, as I mentionned in my note, is that there is a 
>> difference in
>> the determined s values of the same protein in different cells of 
>> AUC. This
>> problem has apparently been seen by others as well. The question is: 
>> what is
>> the cause of the problem? There is no firm answer. There is good
>> speculation.
>>
>> First, I would like to thank you all for your comments. I have 
>> followed your
>> advice. Cell 1 went into rotor position 2 and cell 2 went into rotor
>> position 1. The bias that was found when the cells were in their normal
>> positions (cell 1 in rotor position 1, etc) followed the cells; the 
>> data are
>> clean and clear. In addition to the systematic difference between the s
>> values that I find, there is a systematic difference in the position 
>> of the
>> menisci (meniscuses). Cell 1 which uniformly gives an s value higher 
>> by 0.6%
>> (from
>> stats) has a meniscus that is uniformly closer to the center of the 
>> rotor by
>> about .05 cm. My gut feeling is that this should not in any way 
>> influence
>> the values of s that we determine so long as the meniscus is constant
>> throughout the run. Once I leave the gut, I'm hard pressed to find an
>> explanation for why the value of s should vary. I now apply a correction
>> factor to the s values obtained in each cell.
>>  
>> Last point from a biologist. "I'm learning far more about analytical
>> ultracentrifugation than I ever thought I would want to."
>>
>> All the best and thanks once more
>>
>> jack
>> jack a kornblatt
>>
>>
>> _______________________________________________
>> RASMB mailing list
>> RASMB at rasmb.bbri.org http://rasmb.bbri.org/mailman/listinfo/rasmb
>>
>>
>>
>> _______________________________________________
>> RASMB mailing list
>> RASMB at rasmb.bbri.org
>> http://rasmb.bbri.org/mailman/listinfo/rasmb
>>
>
>

-- 
Department of Biochemistry and Molecular Biology
University of New Hampshire
Rudman Hall 379
46 College Road
Durham, NH 03824-3544
Phone: 603-862-2459
FAX:   603-862-0031
www.camis.unh.edu
www.bitc.unh.edu





More information about the RASMB mailing list