edgeR - DE with one disease sample vs. multiple control samples?

Question

[RafaelMP]

Hello! I am trying to use RNA-seq data to accomplish two objectives in a case vs. control study:

1 - Verify if the expression level is distinct between the case sample and the controls.

2 - We need to identify DE genes between conditions.

Unfortunately, we have only one case with the mutation, but we have six biological replicates for the control group. Is it ok to use the package edgeR, or we need to try something else?

Thanks!

Answer 1

[Gordon Smyth]

There's no problem with this. edgeR handles unequal sample sizes automatically. It will simply estimate variability (i.e., the dispersion parameter) from the controls, where you do have replicates, and will then apply the same dispersion values to both cases and controls when conducting tests and computing fold changes.

This is a frequently asked question:

• Differential expression: replicates in one condition, no replicates in the other
• https://support.bioconductor.org/p/113521/
• https://support.bioconductor.org/p/76631/
• https://support.bioconductor.org/p/63585/
• https://support.bioconductor.org/p/61904/
• https://support.bioconductor.org/p/53627/