Similarity Matrix

http://www.biomedcentral.com/1471-2164/8/353/figure/F5?highres=y
A similarity matrix does exactly what its name indicates: it uses colors to show whether (or not) data sets are similar to each other. Typically, darker colors are used to show a higher similarity between variables, while lighter colors are used to show just the opposite.
This similarity matrix shows the similarities between certain genes. Contrast pairs with significant similarity are coded dark red, while pairs with non-significant similarly have no coloring.