data representation

Mark A. KuKuruga kukuru at umich.edu
Mon Aug 10 13:23:25 EST 1998


Jacqueline Saleh wrote:

> Dear Flow-ers:
>
> Various users at my company are having very heated discussions on what number
> to use to beat reprent fluorescent intensity. Some users are using mean of
> M1(% positive cells) to describe fluorescent intensity. I believe this not is
> inaccurate because
> your using the mean of 20-30% of the cell population and I believe the mean
> of the entire population (not just M1) to be a much more accurate
> determination of fluorescence intensity.

This response assumes you're imparting BD's first stat region (default name:
"M1") with some enhanced level of universal significance (for flow-ists,
anyway).  However, more to the point . .
If the region M1 limits and defines a discreet "positive" population (i.e.,
completely and significantly separate from "negative"), then it is probably
appropriate to compare mean intensity values FOR COMPARABLE SAMPLES of similar
phenotypic positivity.  If, however, M1 merely divides a single gaussian peak
(or even overlapping gaussians) based on your operator's arbitrary M1
placement, then it isn't.  The latter would certainly lead to an erroneous mean
value, and the overall (full scale) mean would be a better indicator.

> I  myself prefer to use percent positive cells as opposed to mean fluorescent
> intensity to describe my data.

As I stated in an earlier response to this debate, sometimes (like when looking
at cell lines, or other homogeneous populations) mean intensity is the ONLY
meaningful comparison, since ALL should be positive.  Example; comparing CD4
expression for activated vs. quiescent T-Helper cells.  All are positive, but
with activation, CD4 level is increased.  With this kind of analysis . . . if
all cells are not positive, something else is going on.

> I also explained to people that you cannot compare mean fluorescent
> intensities from this weeks experiment versus last weeks experiment because
> the means will vary considerably depending on were the negative control is
> set, if the planets are in line with Jupiter, and the "Flow Gods" see fit to
> smile on you.

Whatever . . . more accurately, one CAN directly compare daily/weekly
experiments (even monthly/yearly, or even with the guy next door) provided one
calibrates their cytometer with a known standard of predictable and stable
fluorescence, like a  standard bead set designed for that very purpose.

> Any
> feedback on how people represent their data is greatly appreciated. Thanks in
> advance to all.
>
> Jackie
> Ariad Pharmaceuticals



--
Mark A. KuKuruga, Managing Director
University of Michigan Core Flow Cytometry
kukuruga at medmail.med.umich.edu





More information about the Cytometry mailing list