Hi there, very good work and I believe this work adds to the field. I have one question that is important. while processing the spike-ins bam files do you remove the duplicates or you treat them as the target? In Henikoff's suggested pipeline the duplicate reads are marked but never removed , due to the inherited Tag5 primers insertion. MNAse, the case for cutANDrun does work in a similar fashion and cuts in specific AT rich areas that are NOT random. Thank you for the preprint