BIGOS at Beadle.Stanford.edu
Fri Aug 7 03:54:02 EST 1998
It seems that your question has been interpreted in two different ways.
1) If I run some cells from my sample through my instrument and collect data for X
number of events, and 50,000 of them are positive for some marker and the rest
negative, does that mean the percent positive is 50,000/X ?
2) If I run some cells from my sample through my instrument and collect data for
50,000 events above the threshold level I have set, does that mean I have run only
50,000 cells through the instrument while I was collecting data ?
If (1) is the case, assuming you have a well-prepared sample with little debris and
doublets and an instrument that aborts data points for multiple threshold crossings
during an event measurement period, sampling statistics dominate and instrument dead
time is irrelevant.
If (2) is the case then you need to take into account dead time in order to estimate
the true number of cells that were run during the measurement period. Poisson
statistics will provide a reasonable way to apporximate this, but I believe data has
been published showing arrival times in flow cytometry of mixtures of stained and
unstained cells do not necessarily follow a poisson distribution. Other properties
of the sample, such as clumping, of course will also affect this result.
Stanford Shared FACS Facility
On Tuesday, August 4, 1998 David Burke asked the question:
I recall some time ago that the number of positive events detected by
flow cytometry does not directly correlate with actual number of cells.
For instance, if 50,000 positive events are detected, can one accurately
take that as 50,000 CELLS?
If not, what exactly is the deviation and how can the degree of
deviation be calculated???
Any references or helpful inputs????
More information about the Cytometry