Assembling full-length rRNA genes from short-read metagenomic sequence datasets using EMIRGE

Microbial ecologists have reaped enormous benefit from advances in high-throughput DNA sequencing. However, the short read lengths of currently dominant technologies have made a seemingly simple question about shotgun metagenomic experiments difficult to answer: what small subunit ribosomal RNA (SSU rRNA) genes are present in a sequenced biological sample? Without these gene sequences, it is difficult to interpret a sample within the rich context of ribosomal rRNA databases accumulated over decades. This chapter presents specialized software, EMIRGE, for the assembly of SSU rRNA genes. EMIRGE is optimized to deal with strain similarity and the fluctuating levels of conservation within the SSU rRNA gene that make assembly difficult. It has been used to successfully assemble genes from shotgun metagenomes, long PCR amplicons, and total-RNA transcriptomes. A detailed discussion of how EMIRGE works and how it deals with the uncertainty inherent in the assembly problem is presented. Practical suggestions are given for understanding and optimizing parameter choice, data preprocessing and postprocessing, and creation of a candidate SSU rRNA gene database. When high-throughput sequencing data are available, EMIRGE can serve as a valuable tool for interpreting microbial community structure.