INTRODUCING EzTaxon-e: 16S RRNA SEQUENCE DATABASE WITH NEW HIERARCHICAL CLASSIFICATION SYSTEM FOR BACTERIA

초록

For last few years, DNA sequencing-based community analysis has been widely propagated due to the invention of pyrosequencing. The major challenge in applying pyrosequencing to microbial community analysis lies on dealing with massive sequences and comparing community datasets. The existing 16S rRNA databases have provided taxonomic framework for the identification of bacteria, but inaccurate and insufficient taxonomy on "uncultured bacterium" have hampered correct identification. Here we present the EzTaxon-e, a new 16S rRNA database with complete hierarchical classification scheme for bacteria, which isan extension of EzTaxon database (http://www.eztaxon.org). In EzTaxon-e, we selected representative sequences from public domain to serve tentative type strains of yet uncultivated species using 3% sequence similarity for the species boundary. The classification for uncultured sequences was solely based on extensive phylogenetic analysis and sequence similarity values, resulting in complete and accurate hierarchical taxonomic structure for all species. The new system consists of 88 bacterial phyla and over 25,000 species including 132 candidatus species and >8400 species with valid names. Multiple metagenomic samples from diverse environments were examined to evaluate usefulness of our database. Bacterial community structures revealed using EzTaxon-e database were compared with those based on taxonomic schemes of the Ribosomal Database Project, NCBI and Greengenes databases. Even though community structures were generally similar at the phylum level, our system was shown to have more coverage in the identification of sequences at the family, genus and species levels, especially for as-yet uncultured taxa. We conclude that EzTaxon-e database should provide a sound taxonomic backbone in the identification of massive sequence data and can facilitate comparison of multiple bacterial communities. The database is available at http://www.eztaxon-e.org.