Cell type deconvolution from Bulk RNA-seq data

Question

[Anand]

Hi everyone.

I have brain RNA-seq data and a number of significantly DEGs. I also have a list of marker genes for astrocytes, neurons, microglia, oligodendrocytes, endothelial cells. I've read in some papers that they use a one-sided fisher's test to do the cell type deconvolution analysis to find out which cell types are more affected than others. Does anyone have any code they could share or advice on how to do this?

Thanks.

[asalimih]

Take a look at the EPIC tool . it may help

[Anand]

Thanks. I'll give it a go.

Answer 1

[bigomics.team]

Sound like just doing a Fisher's exact test for geneset enrichment. Most cell type deconvolution methods use the whole expression matrix/vector for sample and references.

Answer 2

[marcin.bazyliszek]

Hi @ Anand

Was this working for you for for astrocytes, neurons, microglia, oligodendrocytes, endothelial cells? I see the tools need some gene expression reference profiles. I would like to do some cell deconvolutions the same fellows.

Cheers, Marcin