Data Display

Maryalice Stetler-Stevenson (stetler@box-s.nih.gov)
Wed, 1 Oct 1997 16:04:03 -0400

I can't resist a comment. After all the discussion on the list, we began
discussing this in the lab. All methods of data display have their use in
specific situations. By being too rigid, one doesn't utilize the best tool.
If you are looking for rare or uncommon events well separated from other
populations, dot plots are very useful. An excellent example is hairy cell
leukemia involving the peripheral blood- we see rare events that are
intensely CD 20 + and CD103 + as well as monoclonal. These would be lost
with a contour plot. If you are looking for patterns within large
populations (eg heterogeniety in levels of antigen expresion in a full
blown leukemia) then contour plots can be great. We shouldn't argue one or
the other but instead use both as needed. As for publication- I'm convinced
no one outside of the flow community can understand flow data figures and
the most important thing is to summarize data so they don't even have to
look at the pictures.

Maryalice

Maryalice Stetler-Stevenson
Director Flow Cytometry Unit
Laboratory of Pathology, NCI, NIH