Table 2.

Supported OTU picking methods in QIIME, with a brief description of the algorithm employed and in which OTU picking approach can be used.

Method	Picking approach			Description	Reference
	denovo	closed reference	open reference
cd-hit	yes	-	-	Applies a “longest-sequence-first list removal algorithm” to cluster sequences.	Li and Godzik (2006) Li, Jaroszewski, and Godzik (2001)
Mothur	yes	-	-	Takes an aligned set of sequences and clusters them using a nearest-neighbor, furthest-neighbor or average-neighbor algorithm.	Schloss et al. (2009)
prefix/suffix	yes	-	-	Clusters sequences which are identical in their first and/or last bases.	QIIME team, unpublished
Trie	yes	-	-	Clusters sequences which are identical sequences and sequences which are subsequences of other sequences.	QIIME team, unpublished
blast	-	yes	-	Compares and clusters each sequence against a reference database of sequences.	Altschul et al. (1990)
uclust	yes	yes	yes	Creates seed sequences which generate clusters based on percent identity.	Edgar (2010)
usearch	yes	yes	yes	Creates seed sequences which generate clusters based on percent identity, filters low abundance clusters and performs de novo and reference based chimera detection.	Edgar (2010)