on June 21, 2011 by Daniel Swan in Reviews, Comments (1)

Review of “Volcano plots of microarray data”

This is a review for “Volcano plots of microarray data” by Colin Gillespie.

The tutorial follows on from “Analysing Microarray data in BioConductor” and shows a widely-used visualisation technique called a volcano plot, which allows negative log10 p-values to be plotted against log fold change between two experimental conditions.

I think it’s worth noting that the dataset in question has a large number of significantly differentially expressed genes between the cell line and the endothelial tissues when analysed, but limma provides adj.P.val so whilst the Bonferroni approxomiation is a nice technique for the plot, would it make more sense to use the adj.P.Val for a volcano plot rather than taking the unadjusted p-value from topTable?

It’s really good to highlight just how much better ggplot2 is for producing publication quality figures, something which is worth repeating.

I’m curious if it is possible though to label individual points (or a small subset of) in the plot so that a gene name or Affymetrix identifier might be overlaid near to its graphed position?

1 Comment

  1. Colin Gillespie

    June 21, 2011 @ 3:32 pm

    Thanks for the review. You make some very good points which I have now addressed in the article.

Leave a comment

Login