I used the uchime-denovo command from the vsearch plugin in a recent workflow and it identified additional chimeras beyond DADA2. Nice. However, I ran into a hiccup when exporting nonchimeras.qza to a fasta file in that the sequences were broken across 80 character lines instead of using a single line. Not a big problem, but it would be nice if the format was either set to '--fasta_width 0' to be consistent with DADA2 or inherited from the input .qza.
Thanks!
Chris