How to read a volcano plot
A volcano plot shows every gene from a differential expression analysis as one point, so you can see at a glance which genes changed a lot and which changes are statistically convincing.
What the two axes mean
The horizontal axis is the log2 fold change: how much a gene's expression differs between two conditions. A value of 1 means the gene doubled, 2 means it quadrupled, and -1 means it halved. Zero means no change.
The vertical axis is -log10 of the adjusted p-value. The transformation flips the scale so that stronger evidence sits higher. An adjusted p-value of 0.05 becomes 1.3, 0.001 becomes 3, and so on.
How to interpret the regions
- Upper right: genes that increased substantially, with strong statistical support. These are your up-regulated genes.
- Upper left: genes that decreased substantially, with strong support. These are your down-regulated genes.
- Bottom: genes whose change could plausibly be noise, whatever its size.
- Top centre: genes with a small but very consistent change. They are statistically significant, though the effect may be too small to matter biologically.
The plot gets its name from this shape: a dense base of unchanged genes with two plumes rising on either side.
The cutoff lines
Most plots draw a horizontal line at the significance cutoff and two vertical lines at the fold change cutoff. Common choices are an adjusted p-value below 0.05 and an absolute log2 fold change of at least 1. These are conventions, not laws. A stricter cutoff gives a shorter, more reliable gene list, and a looser one gives a longer list with more false positives.
Common mistakes
- Using the raw p-value. With 20,000 genes tested, a raw cutoff of 0.05 yields about 1,000 false positives by chance. Plot the adjusted p-value (called
padj,FDRoradj.P.Val, depending on the software). - Reading the plot as a ranking of importance. The most extreme point is the most statistically striking gene, not necessarily the most biologically important.
- Ignoring expression level. A gene with very few reads can show a huge fold change from a handful of counts. If a surprising gene tops the plot, check its average expression.
- A flat ceiling of points. Some software reports p-values of exactly zero below a numerical limit, which lines the top genes up in a row. It is a display artefact.
Try it on your own results. The DE Results tool on ExpressionBench turns one results table into a volcano plot, a gene table and pathway enrichment. It is free, and your data stays in your browser.
Open the tool