Re: Signal detection and sensitivity

From: Christopher S Boyce (csboyce@beckman.com)
Date: Thu Jun 01 2000 - 20:11:30 EST


Hello Robert,

Check out the new journal article in this month's Cytometry:

"Polarization of Scatter and Fluorescence Signals in Flow Cytometry"
Volume 40 Number 2, June 1, 2000.
This might address  your problem.

I have found that in air-in-stream sorters/analyzers, slight adjustments to  the
Z-axis, scatter bars,  and slight adjustments
to the laser beam path, laser power, and the FL focus, CAN help you get signals
that look like they do on another instrument.  But this is an art.  However, it
can be qualified using bead comparisons, or CRBC comparisons between the two
instruments, comparing channel numbers and scatter patterns using the same
filters.

I had a similar problem with a FACSCAN vs. FACSTARplus.  My supervisor was not
satisfied until the scatter and FL signals were nearly identical. Well, this is
nearly impossible, without having both instruments in your hands, and working
with ALL the variables in setting up a good interogation point on a
stream-in-air system to mimic what is seen in a cuvette flow chamber.   I was
able to closely match the  same signals, and use settings appropriate for each
instrument to get similar results. I used beads of many types, and  CRBC's and
tried to "mimic" the signal produced by the FACSCAN.  This worked well, but it
was not exact.  I am sure there are better, published, ways to do this.

It is a given that without putting the screws to both instruments, the plots
may not look the same, but the analysis-end result should be similar if the
experiment is designed well (multiple controls, right filters, etc.), and  you
accept the slight differences between the two instruments.


Howard Shapiro has an interesting editorial about the article I mentioned above
in the same journal.  He states-- in general, polaraization of the laser beam
and how it interacts with the particles, filters, lenses, and mirrors affects
end results.  And not all instruments polarize alike!  Yikes.  The above
article lends credence to my practice of telling  investigators not to expect
the results to be exactly the same on different instruments (but this is not an
excuse for a misaligned, or otherwise poorly working instrument).  And I suggest
using one instrument throughout an experiment's long term execution.  Also, when
going from an analyzer to a sorter---for the purpose of sorting, ( after much
analysis of what to sort on the analyzer), a problem arises when investigators
are not  seeing plots exactly like they have in their hands.  And guess who gets
the blame?

Granted, making sure you have the correct filters (or improved filters)  is
always a good idea.





Jens Fleischer <jfleischer@knuut.de> on 05/31/2000 06:48:22 PM

Please respond to jfleischer@knuut.de

To: cyto-inbox
cc:    (bcc: Christopher S Boyce/BII)

Subject:  Re: Signal detection and sensitivity



Robert Adair McGilp wrote:

> I have an investigator who has been using a FACSort to analyse his FITC/
> PE/Cychrome labelled samples and gets good resolution of the populations
> of interest (dual positives). He feels that we have been unable to
> reproduce the results on our Vantage SE. We have moved the z axis so that
> the nozzle is a good inch above the laser intercept and this gives a good
> approximation of the results he has seen on the analyser.

Be sure to insert the appropriate filter in front of the FL3 detector.
Pe-Cy5 has its emission maximum at 675 nm. BD has a 675/40 bandpass
filter and a 650 longpass. Both should give a good resolution. In the
standard configuration you often find a 630/22 or 610/20 designed for
propidium iodide in front of FL3, which will not give good intensities
for PE-Cy5.

Jens



This archive was generated by hypermail 2b29 : Sat Mar 10 2001 - 19:31:22 EST