(Flow) Data presentation 1 - Histograms

I would say the easiest thing about flow cytometry is the acquisition of data, most flow cytometers are easy to use and thousands of events can be recorded in a short space of time. We must however understand how to interpret this data and importantly how to present it in talks, on posters or in publications. There are a number of ways of doing this, but they essentially revolve around the two types of charts we generate on a flow cytometer, namely histograms and dot plots.

In this series of blogs I would like to convey a few ideas I have about data presentation. A good resource for this is Roederer et al (2004) Guidelines for the Presentation of Flow Cytometric Data. Methods in Cell Biology, 75:241-257.

So to begin with we'll look at histograms.

Histograms are maybe the easiest chart for us to use to display our data. They comprise of a single parameter plotted against frequency. Often this parameter is a fluorescent signal and in most cases comprises a log scale to enable low and high intensity events to be displayed on the same chart. The exception to this would be cell cycle analysis where we are looking to determine the relative amounts of DNA in cells to ascertain where they are in G1 or G2 or the relevant numbers of chromosomes present.

Histograms can be presented in a number of ways however. Firstly we can simply show a number of histograms that represent each of our experiments. This can be a little laborious to look at and any subtle changes in data between experiments would be lost. A solution to this would be to produce overlays where we can plot our control population and our experimental data sets on the same histogram. As an example we could format the histogram so that the negative population was displayed as a line and the test data superimposed on this as filled in areas (see figure). This way the important test data is not overpowered by the control data. Along the same lines we could use a low contrast colour for our control sample and use distinctive colours for our test data.

http://jcsmr.anu.edu.au/facslab/analysis.html

A third option is to present our histograms in a three dimensional format. This would be useful to show a trend over a period of time or a series of drug treatments for example. This following example is from the Purdue University Cytometry Laboratories (http://www.cyto.purdue.edu/flowcyt/research/micrflow/jepras/jepras2.htm)

Well that's a start to data presentation of flow cytometry data, I'll discuss dot plots next time!