SSUnique: Detecting Sequence Novelty in Microbiome Surveys

• Published in: mSystems
• Main Authors: Michael D. J. Lynch, Josh D. Neufeld
• Format: Article in Journal/Newspaper
• Language: English
• Published: American Society for Microbiology 2016
• Subjects: 16S rRNA; high-throughput sequencing; microbial dark matter; microbiome; rare biosphere; taxonomic blind spots; Microbiology; QR1-502; Arctic; Tundra
• Online Access: https://doi.org/10.1128/mSystems.00133-16; https://doaj.org/article/ec8357535c4b40f2b4f0b8af56646b1b

ABSTRACT High-throughput sequencing of small-subunit (SSU) rRNA genes has revolutionized understanding of microbial communities and facilitated investigations into ecological dynamics at unprecedented scales. Such extensive SSU rRNA gene sequence libraries, constructed from DNA extracts of environmental or host-associated samples, often contain a substantial proportion of unclassified sequences, many representing organisms with novel taxonomy (taxonomic “blind spots”) and potentially unique ecology. Indeed, these novel taxonomic lineages are associated with so-called microbial “dark matter,” which is the genomic potential of these lineages. Unfortunately, characterization beyond “unclassified” is challenging due to relatively short read lengths and large data set sizes. Here we demonstrate how mining of phylogenetically novel sequences from microbial ecosystems can be automated using SSUnique, a software pipeline that filters unclassified and/or rare operational taxonomic units (OTUs) from 16S rRNA gene sequence libraries by screening against consensus structural models for SSU rRNA. Phylogenetic position is inferred against a reference data set, and additional characterization of novel clades is also included, such as targeted probe/primer design and mining of assembled metagenomes for genomic context. We show how SSUnique reproduced a previous analysis of phylogenetic novelty from an Arctic tundra soil and demonstrate the recovery of highly novel clades from data sets associated with both the Earth Microbiome Project (EMP) and Human Microbiome Project (HMP). We anticipate that SSUnique will add to the expanding computational toolbox supporting high-throughput sequencing approaches for the study of microbial ecology and phylogeny. IMPORTANCE Extensive SSU rRNA gene sequence libraries, constructed from DNA extracts of environmental or host-associated samples, often contain many unclassified sequences, many representing organisms with novel taxonomy (taxonomic “blind spots”) and potentially unique ecology. This ...