As I’ve come across more experimental designs, I have come to question the “universality” of the approach demonstrated in this tutorial. In particular, for partially crossed imbalanced designs (see issue #9), where this approach would in fact create a wrong contrast vector. When (if) I have time to revise this tutorial, I will come back to it. Until them, please see these instructions as personal notes from someone who is still learning and may get things wrong.
A general recommendation when working with more complex designs and custom contrasts is to always check the results output against the original data.
For example, plot the original data points for a given “significant” gene, to see if the expression in the individual samples makes sense with the logFoldChange reported from the model output.
If there’s a big discrepancy, then it’s possible the contrast was not set correctly.
Despite the non-universality of this approach, I think the tutorial still gives a reasonable intuition of how contrast vectors work with DESeq2, at least for relatively simpler designs.
This repository contains some code explaining how to set DESeq2 contrasts using the model matrix of our design.
This approach is general and so works for a range of experimental designs, even more complex ones. (see warning above)
If something doesn't look correct, please let me know (open an issue). I would love to learn if I've gotten something wrong!
Using these materials
Feel free to use and adapt these materials to your own needs. If you do, please give attribution to: Hugo Tavares (University of Cambridge, Bioinformatics Training Facility)