Same results in Limma Toptable as in DESEQ2 results

Question

[Bine]

Dear all,

I am trying to generate exactly the same results in Limma for my result table as I did in DESEQ2 (Differential Expression Analysis).

With the toptable function I am not getting lfcSE and basemean.

Can anyone help me how to extract this information via Limma?

I do get logFC, AveExpr, t, P-Value, adj.Pvalue and B in Limma.

Thank you, Bine

[jared.andrews07]

These are different tools with slightly different approaches. Why would you expect the exact same results? AveExpr is the limma equivalent of DESeq2's basemean column, but it has no lfcSE equivalent.

[Bine]

Thank you very much. I am replicating the analysis with Limma and for my result tables I would like it to be exactly the same output.

[jared.andrews07]

I get that. My question is why would you expect that to be plausible? These methods account for similar things and tend to return largely overlapping results, but they have differences in approach and thus results as well. You're not going to get the exact same results, and I don't know why you'd expect to. Robust findings should be reasonably replicable between the major DE methods (DESeq2, edgeR, limma).

Answer 1

[Gordon Smyth]

It would be superfluous to include a SE column in the limma output because it is not needed and because it is immediately computable from the other columns.

[Bine]

Ok thank you. Following this t: logFC divided by its standard error I should get the SE by SE=logFC/t or?

[Gordon Smyth]

Yes.