Re: cytometry of cyclins

From: Mark A. KuKuruga (kukuru@umich.edu)
Date: Wed Mar 31 1999 - 15:07:51 EST


Lindsay Robinson wrote:

> Hi,
> I have noticed in the literature (eg. Darzynkiewicz et al. Cytometry 25:1-13,
> 1996) that cyclin/DNA data is commonly analyzed by contour plots. I was just
> wondering if there is a reason for using contour plots as opposed to other
> methods such as dot plots or histograms (I realize histograms would be for
> cyclin expression alone, not with PI).

Lindsay,Two things here . . . 1) Contours offer more control over the display
of the data, and will often provide a better, clearer indication of the areas
of concentration in a distribution by limiting the display of satellites.  The
data, however, (stats-wise) is the same in a dot plot as in a contour plot.  2)
You can use single parameter histograms to display and analyze these data, but
often the correlation with cell cycle distribution is the whole point, so
correlation with DNA content is essential.

> In our lab we have done (up until now) mostly lymphocyte work and we are used
> to analyzing our immune cell data by dot plots when we do double and triple
> labelling. Can I analyze my cyclin/DNA data using dot plots? I have tried it
> using FL2-A (PI stain) versus FL1-H (FITC labelled cyclins). When I use dot
> plots I am having trouble knowing where to set the quadrant...S phase has to
> be either in the upper or lower quadrant and I don't know if this is correct

You can use quadrants as described.  You use the isotype control to establish
background, then run your positives against that quadrant
setting.Alternatively, you can use regions (parallelograms), setting the region
again above the isotype, and look for positive events that way.  This is
sometimes useful, since often the background for later DNA fractions (G2, M,
late S-phase) can be different -- higher -- than the earlier (G1, early S).
So, the parallelogram allows for this by accommodating a "slant" in he
distribution.  Sometimes, because the quadrant method can only accommodate
"flat" regions, this latter method can be more accurate.

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



This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:53:19 EST