Nice paper on how to measure classification performance on 16S and ITS data. Things haven't improved much since 2007. Best hit is usually right, the challenge is what taxonomic level to stop at. https://t.co/XgH3UaNyAP
@bebatut Some good points but with of course commercial bias in there. Edgar's most recent paper is a challenging read (and doesn't only cover Greengenes): https://t.co/hysAzHQRhj