This is a test version of Biostars. For the public version, visit https://www.biostars.org.
DEseq2 analysis for RNAseq data, possible to do pair-wise comparisons?

Hi all,

I'm currently analysing some RNAseq data and my colleague has asked about pair wise comparisons. I'm still very new to bioinformatic analysis, having only started a couple of months ago on it, so wanted to seek some advice and guidance.

In the study, he has 3 samples of 2 conditions (wild type, and mutant). The RNA from each paired set of samples (1 WT and 1 mutant) were extracted on the same day, and the different pairs on different days.

The RNAseq was performed, and they seem to cluster in their respective pairs, but with some (more than i'd want to see) variability between the pairs. Is it possible to perform differential expression to take into account these paired samples, I've done DEseq on the 6 samples all together. Thanks in advance

rna-seq deseq2
• 1,240 views • link
updated by rpolicastro rpolicastro • written by dnljmrs • dnljmrs

1 answer

Thankfully since the batches contain samples of each type, you can add batch as a covariate in the regression model. First, add a column in your meta-data table that contains a categorical factor for each batch. Your table would look something like this.

     condition batch
WT1         WT     A
WT2         WT     B
WT3         WT     C
MUT1       MUT     A
MUT2       MUT     B
MUT3       MUT     C

Then, your regression model would be ~ condition + batch. This would return the results of condition while reducing the effect of batch.

As a side note, if you are curious whether there is a statistically significant batch effect, you can run the likelihood ratio test described in the DESeq2 documentation.

• 0 views • link
written by rpolicastro rpolicastro

Thank you so much, I'll give that a go :)

• 0 views • link
written by dnljmrs • dnljmrs

Log in to answer this question.

More posts like this