How do I visualise an expression matrix?
Make a heatmap
Clustered heatmaps, and the scaling decision that changes what they show.
Nearly every decision here is about scaling. Row-scaling (z-scores per gene) shows relative patterns across conditions and discards absolute abundance; unscaled data shows abundance and lets highly expressed genes dominate the colour range. Both are legitimate and they answer different questions, so state which you used.
Clustering rows groups genes with similar behaviour. Clustering columns is a check on your experiment: if replicates do not cluster together, that is worth knowing before the figure.
A red–green colour scale is unreadable to roughly 8% of men. Use a perceptually uniform diverging scale such as blue–white–red instead.
What to use
- Morpheus
Interactive heatmaps with clustering, in the browser.
ExternalBroad InstituteFree - SRplot
Free web platform covering a very wide range of biological plot types.
ExternalScience and Research Online PlatformFree - R and RStudio
The statistical environment nearly all biological methods are written for.
ExternalR Foundation / PositFree
If you script
Command-line and library options. No page here — they need an environment rather than a browser.
- ComplexHeatmapR / Bioconductor package
The reference R package for annotated, publication-grade heatmaps.