Hi Aura,

You may run the differential expression analysis between your cell cluster of interest and other cells in the dataset. From there you will get a list of genes, enriched in your cell cluster of interest. Then you feed this list to Kobas (or similar) for gene set enrichment / gene ontology analysis. The enriched GO/Reactome/PANTHER/BioCyc/Patway terms may give you a hint about what cell type it is.

Best,

Michail

Hi Michail,

Thanks for the suggestion!I will try and see if it helps!

Best,

Aura

Hi Aura,

cluster identification can be very challenging, especially if it comes to previously unknown cell types. Probably the best solution is to have an experienced biologist with good knowledge of genes have a look a the topmarkers (top DE genes of the cluster of interest).

As mentioned the enrichment analysis can also help you here. I highly recommend enrichr ( https://amp.pharm.mssm.edu/Enrichr3/), which is also very nicely integrated in R ( https://cran.r-project.org/web/packages/enrichR/vignettes/enrichR.html ).

You could also use SingleR ( https://github.com/dviraran/SingleR ) which tries to automatically identify cluster based on reference dataset (my experience is that it is very good at identifying basic cell types like T cells, B cells, but struggles when it comes to rarer cell types).

Best,

Mischko

Thanks Michael Heming ,

I will try these methods as well! Most of my cells are not T and B, so I will go for the enrichment analysis and see how that goes.

Best,

Aura