You could use image() and then text() , but keep in mind you may need to transform and reverse the row order to get it to look like your data frame. Also, you'll have to make the scale separately. Here's an example for the heatmap part... Or here's a messy script that I've used.

You could try pheatmap also.

Seems similar to what you can produce with heatmap.2, except I don't typically see values on each cell (especially if aren't necessarily used for clustering/visualization, such as the p-values). It is an interesting alternative to the venn diagrams in Figure 2 (which I think you can produce using the vennerable package).

Perhaps it is some custom implementation of the default heatmap() package?

http://flowingdata.com/2010/01/21/how-to-make-a-heatmap-a-quick-and-easy-solution/