Appropriate Use of Overlay Gates

FlowJo provides the ability to overlay one population on top of another, even if you are using a bivariate plot (e.g. CD4 vs CD8, as opposed to a histogram. We've been using this extensively to present cytokine producing T cells. However, it must be done carefully in order not to mislead you and your audience.

Basic Rule of Thumb

An example

We've been looking at the tropism of VSV using a recombinant expressing GFP. In the past, we showed that the VSV.GFP infects all monocytes, but not T cells and B cells. We decided to look more carefully at monocyte subsets.

We set up a basic four-color flow panel for a human whole blood stain as follows:

Gating on Lin- population

First, let's consider the gating strategy. Here's what we did originally:

  1. Plot FSC vs SSC and gate on all leukocytes, including the granulocytes.
  2. Plot the leukocyte population as Lin vs HLA-DR, and set a rectangular gate on all Lin- cells.

    As we'll see below, there was probably a better way of doing this.

  3. Finally, for cells in the Leukocyte / Lin- population, plot HLA-DR vs. CD11c.
  4. For the GFP+ gate, cells in the leukocyte population were plotted as FSC vs. GFP, and a rectangular gate was set on GFP+

    As we'll see below, this was fine as far as it went, but it should have gone further. But I'm getting ahead of myself.

The left-hand panel of the figure below shows how the gates were set (except for GFP), while the other panels present alternative approaches. I prefer the one in the middle, because I don't like to gate on a parameter and than use that again in the next subset. In the example below, it also gave me another chance to gate out more FSC-low debris that slipped through my original leukocyte gate.

The overlay plots

Here's how we generated the overlay plots. The left-hand panels in the figure below get us our Lin- leukocyte population.

In the third panel above, the red contour plots represent all cells falling within the Leukocyte and Lin- gates, while the blue dots represent cells falling in a Leukocyte / GFP gate. This strategy is incorrect, because the GFP gate should have been set on the Lin- population. The correct plot is shown at the far right, and presents a blue overlay on top of a red dot plot, representing all Lin- leukocytes. Where are the blue dots? There aren't any, because there were not GFP cells among the Lin- cells.