Hello,
In the usearch program, it suggests not generally doing any additional chimera detection:
"The chimera filter in cluster_otus is very effective, so I generally don't
do any additional filtering.
It may be possible to find some additional chimeras by running
uchime2_ref on the OTU sequences generated
by cluster_otus. You should use
‑mode high_confidence to reduce the risk of false
positives." - UPARSE pipeline instructions online.
Since (to my knowledge) there isn't an identical cluster_otus program (instead there seems to be --cluster_fast, --cluster_size or --cluster_smallmem) is the clustering function in vsearch the same? Are additional chimera detection steps also not generally recommended for vsearch?
Thanks,
Julia