Hi @Daniel Lundin I talked with Alexey yesterday, the main author of SATIVA, at present he does not have time to look at this and thinks that in principle the entire approach should be ported to using EPA-NG (academic.oup.com/sysbio/…/5079844). However, EPA-NG was developed by a former PhD student of mine who now works in industry. Nontheless, we could ask my other former PhD student Lucas Czech who fortunately remained in Academia and also worked on phylogenetic plecement if he would be interested. We may also consider Giorgo's idea to use raxtax for some prefiltering and subsequently run SATIVA on a much smaller pre-filtered dataset