Using findmarker of seurat with different test.use but getting the same resulat.

Question

[Sep]

Dear all,

I have a seurat object with 15292 features and 731441 cells. I use the following functions:

possion= FindMarkers(object = combined_pbmc, 
            ident.1 = "YES",  #
            ident.2 = "NO", 
            test.use = "possion",
            assay = "RNA",
            slot = "counts"
            )

_Negbinom= FindMarkers(object = combined_pbmc, 
            ident.1 = "YES",  # or whatever your disease status categories are
            ident.2 = "NO", 
            test.use = "wilcox",
            assay = "RNA",
            slot = "data"
            )

The differential expression analysis between the case and control group, Although different test.use is used in each function, giving the same number of selected genes and gene names but the order is a little different. does it make sense or there is sth wrong?

[bk11]

Unless you specify the different sets of genes in your two different test, you ll get the result for the same gene list. However, there will be slight shuffle in P and FoldChanges values for each genes from two tests. Bottomline: the result your are seeing totally makes sense and is expected.

Answer 1

[yura.grabovska]

So it's not giving an identical result? Are the fold-changes and p-values different? If it's just giving the same number of variable genes but isn't actually giving an identical result, then probably there isn't an error rather you're just working with a fairly robust dataset where the result isn't as affected by the method used.