"Mislabeled SRA entry is one thing but - it’s clear that it’s impossible (in my hands at least with a -very good- pipeline) to assemble the reference that is in GenBank from the data in SRA"