Hi,

I was trying to analyze a dataset with a slightly high number of duplicates, (i.e. the duplication rate at the lower end of the expression (reads/kbp) seems to be high at ~ 25%). I am therefore not too confident in the results derived at this end of the expression spectrum. Would it be reasonable to use 'basemean' to filter out results that have low counts? Else, how should I proceed?

many thanks!

Yes you can filter on baseMean if you like -- I don't know about the advisability wrt duplicates, but DESeq2 already filters on baseMean during results.