Standards for Sequencing Viral Genomes in the Era of High-Throughput Sequencing

Ladner, Jason T.; Beitzel, Brett; Chain, Patrick S. G.; Davenport, Matthew G.; Donaldson, Eric; Frieman, Matthew; Kugelman, Jeffrey; Kuhn, Jens H.; O’Rear, Jules; Sabeti, Pardis C.; Wentworth, David E.; Wiley, Michael R.; Yu, Guo-Yun; Sozhamannan, Shanmuga; Bradburne, Christopher; Palacios, Gustavo

Population-level characterization. HT sequencing technologies provide powerful platforms for investigating the genetic diversity within viral populations, which is integral to our understanding of viral evolution and pathogenesis (10, 11) . Population-level characterization requires very high levels of HT sequencing coverage (12, 13); however, the exact level will depend on the background error profiles of the sequencing technology and the desired level of sensitivity. As an example, Wang et al. (12) determined that for pyrosequencing data,~400ϫ coverage is necessary to identify minor variants present at 1% frequency with 99.999% confidence, and~1,000ϫ coverage is needed for variants with a frequency of 0.5%. Targeted amplification of the viral genome is often necessary to achieve these coverage requirements. Due to the modest sequence lengths of most HT technologies, the state of the art for population-level analysis has been the characterization of unphased polymorphisms. However, single-molecule technologies, with maximum read lengths of Ͼ20 kb, are opening the door for complete genome haplotype phasing (14) .