Hi there,

I tried to re-analyse RNA-seq data that was made available on GEO and referred to in a publication.

However, the only "raw" data that was made available was the rlog table extracted from assay(rld) (at least that's what I understand from the labeling of the provided table).

Is there any way to feed this table with rlog values into the DESeq function and defining the groups I want to compare?

Or would I need the raw count table to actually re-do the analysis? I would assume so, but I wanted to double-check.

Many thanks

Saskia

You need the raw counts to exactly reproduce the DESeq2 analysis. There is no easy way back from the rlog counts.

You could try ordinary t-tests, linear models or limma on the rlog-data. The results should be similar (depending on data quality), but will not be the same.
 

Thanks Wolfgang, I thought so, I just wanted to make sure before I might mention to them that they did not provide raw data and therefore make it impossible to reproduce their results.

Cheers

Saskia