Deseq2

Question

[odi]

Background: I have my count matrix (with only 4 columns: gene, meso1, meso2, meso3) and it needs to be compared to publicly available mesothelial samples from patients to account for how accurate they are. This publicly available data is already normalized and has 8 columns: gene, meso1, meso2, meso3, mesomd1, mesomd2, mesomd3, and mesomd4.

My count matrix is not normalized. How would you recommend that I combine this data in order to carry out deseq2 analysis and compare your count matrix to publicly available mesothelial samples while also analyzing the differential expression between (control, and HGSOC)?

meso1-3 (are the controls)

[bk11]

Please remember that you cannot use already normalized data in DESeq2 analysis, you need a raw count matrix. Having said that if you have raw count matrices for your and publicly available data, you could combine them perform DE. But, you should be aware that various factors (for e.g, batch effect etc. etc.) will come into play to affect your results.

Answer 1

[swbarnes2]

These are the exact same samples, done at two different sites?

They will be different. RNASeq is very sensitive to batch effect. You generally can't compare samples done by one lab with samples done by another.