Diffbind with <3 reps?

Question

[a_bis]

Hi all, I'm trying to carry out a diffbind analysis through RStudio on some ATAC-seq data to determine differentially accessible sites between two conditions. Unfortunately, I only have two replicates per condition.

When trying to analyse the data on RStudio, I get a message saying "No contrasts added. There must be at least two sample groups with at least three replicates." Conversely, I seem to be able to carry out a diffbind analysis with these datasets on the galaxy platform. For the sake of consistency with previous work though, I would like to carry it out through RStudio. Is this possible? Thanks!

Answer 1

[Rory Stark]

I'm not sure what version of DiffBind you are using, or what your code looks like, but you can override the three-replicate requirement by calling dba.contrast() directly with minMembers=2.

[a_bis]

Thank you, this worked!