DE Analysis on Scran-Normalized Counts

Question

[lucia.liu54]

I would like to perform differential expression analysis on a scRNA-sequenced GEO dataset (GSE111065). I can use either raw counts from featureCounts, or normalized counts prepared via the deconvolution method in scran. I would like to use the normalized counts if possible, but am unsure if I can use them as input into programs like Seurat. I'm interested in figuring our how a few very specific genes change between cells of different types.

Answer 1

[jared.andrews07]

If you're already using scran, why don't you just use Bioconductor packages for your DE as described in the OSCA book? Chapters 11 and 14 may be of particular interest to you.

[lucia.liu54]

Thank you! I'll look into those chapters.