Secondary structure improves OTU assignments of 16S rRNA gene sequences

In contrast to the commentary of Wang et al. (2011) , I contend that the inclusion of secondary structure information provides a faster and more robust analysis when assigning sequences to operational taxonomic units (OTUs). The authors ignored the long history of molecular phylogenetics, where multiple sequence alignments are preferred to pairwise sequence alignments because multiple sequence alignments preserve positional homology across all sequences, not just pairs of sequences ( Durbin et al., 1998 ). Furthermore, profile-based alignments that incorporate the secondary structure of the 16S rRNA molecule are preferred because they provide additional biological information that strengthens the confidence that positional homology is being conserved ( Keller et al., 2010 ). The degree to which these theoretical points are important is contested. I contend that profile-based alignments using curated secondary structure models combined with hierarchical clustering algorithms should be the standard method of assigning 16S rRNA gene sequences to OTUs. I will justify my position by re-emphasizing my previous work, which the Wang commentary ignored, and illustrate several incorrect conclusions that they reached in supporting their thesis ( Schloss 2009 , 2010 , 2011 ; Schloss and Westcott, 2011 ).