Howard has a good point- I didnt mean in this case to suggest that contour plots would be better. They would simply serve to illustrate where the cells fall in terms of phenotype, and "show up" better than dots. Of course the danger is that contour plots can be misleading, if one doesnt look carefully at the number of cells used to construct the plot. The dot plot could be "helped" with an arrow to direct the readers attention to the "relevant" dots. The right thing in this case might be to go back an collect a much larger number of events, if possible. ======================================================= Rachel M. Gerstein, Ph.D. Department of Molecular Genetics and Microbiology Graduate Program in Immunology/Virology University of Massachusetts Medical School 55 Lake Avenue North Worcester, MA 01655-0002 (508) 856-1044 (508) 856-5920 (FAX) > ---------- > From: Howard Shapiro > Sent: Monday, December 23, 2002 3:08 PM > To: Gerstein, Rachel; Cytometry Mailing List > Subject: RE: help with dotplots > > Rachel Gerstein wrote- > > >use contour plots, but note the number of cells displayed in the plot if > >its a small > >number > > In response to Niga Nawroly's posting: > > > > One of our lab's searchers have been looking at gamma delta T cells (very > > > low percentage of PBMCs). Now, he wants to present his data in dotplots > > > (paper quality). The problem is that, he did not acquire enough cells to be > > > able to see a nice population when it is printed out. > > > > > > My question is to know if there is a way by using a software to make the > > > dots look more realistic (each dot representing one cell or half a cell!). > > > > > > PS: We use winlist, cellquest and WinMDI > > > > > In this instance, contour plots will make the data look prettier, not more > realistic. You can't get more realistic than the dot plot. I think it's > fair to say that some of the most vocal advocates of contour plots favor > overlaying the single occurrences as dots when it is these low-frequency > events which are of significance. But, whether you do the contour plot, the > contour plot with overlaid dots, or the dot plot, you should note the > number(s) of cells represented. > > -Howard > > > >
This archive was generated by hypermail 2b29 : Sun Jan 05 2003 - 19:26:32 EST