TABLE 2.

A summary of highlighted methods and pipelines for microbiome data processing.

Steps	Sub-step descriptions	Highlighted methods and their availability in popular pipelines (QIIME, MOTHUR, and UPARSE)
(1) Quality control	Chimera removal and noise mitigation	Trimmomatic(Q) (Bolger et al., 2014), AmpliconNoise(Q, M) (Bragg et al., 2012), UNOISE(M, U) (Edgar, 2016), UCHIME(Q, M, U) (Edgar et al., 2011), Deblur(Q, M) (Amir et al., 2017), and DADA2(Q) (Callahan et al., 2016)
	Remove host DNA contaminant reads	Bowtie2(Q) (Langmead and Salzberg, 2012), BMTagger (Agarwala and Morgulis, 2011), and DeconSeq (Schmieder and Edwards, 2011)
(2) Sequence assembly	De novo read assembly	MEGAHIT (Li et al., 2015), MAFFT(Q, M) (Katoh and Standley, 2013), UCLUST(Q, U) (Edgar, 2010), and metaSPAdes(Q, M) (Nurk et al., 2017)
	Read alignment to annotated database	DIAMOND (Buchfink et al., 2014), NAST(Q, M) (DeSantis et al., 2006), USEARCH(Q, U) (Edgar, 2010), and VSEARCH(Q, M) (Rognes et al., 2016)
(3) OTU analysis	Assignment of reads to OTUs	UPARSE-OTU(U) (Edgar, 2013), Kraken (Wood and Salzberg, 2014), MetaPhlAn2(Q) (Truong et al., 2015), and DOTUR(M) (Schloss and Handelsman, 2005)
(4) Functional profiling	Functional profiling and prediction	MEGAN (Huson et al., 2016), HUMAnN (Abubucker et al., 2012), MetaCLADE, MOCAT (Kultima et al., 2016), and PICRUSt (Langille et al., 2013)
(5) Diversity analysis	Diversity, evenness, and richness metrics	Alpha [e.g., Chao1(Q, M, U)] and Beta [e.g., Jaccard(Q, M, U)]