Is there a limit for the sequence coverage difference between replicates and among samples?

I am using DESeq2 and edgeR in DiffBind to compare ATAC-seq samples with different sequencing coverage among samples, from 77 to 25 million reads, and up to 4 fold difference in sequencing coverage between biological replicates.

I assume that there is a limit for these differences, and I would like to know if we should expect a warning or an error when exceeded.

Thanks,

Cesar

I am not familiar with the insides of DiffBind, but edgeR does not impose any restrictions on sequencing depths. A four fold variation in depth between samples is not at all a problem. edgeR is routinely used for sequencing depths very much more variable than those you mention. I assume the same is true for DESeq2.