Complete viral RNA genome sequencing of ultra-low copy samples by sequence-independent amplification

Malboeuf, Christine M.; Yang, Xiao; Charlebois, Patrick; Qu, James; Berlin, Aaron M.; Casali, Monica; Pesko, Kendra N.; Boutwell, Christian L.; DeVincenzo, John P.; Ebel, Gregory D.; Allen, Todd M.; Zody, Michael C.; Henn, Matthew R.; Levin, Joshua Z.

In this study, we used a sequence-independent amplification method, coupled with Illumina sequencing to generate full genome assemblies for HIV, RSV and WNV samples with as little as 100 viral RNA genomes. Previous studies using Ovation RNA-Seq with HIV viral RNA used ex vivo amplification, large amounts of viral RNA, and utilized a reference-based assembly approach (39) . For several reasons, these conditions are not ideal for sequencing from clinical samples. First, most clinical samples contain a few picograms or less of viral RNA. In our study, we were able to study viral samples with femtogram and attogram amounts of viral RNA (over 1 million times less material than had been previously used) as input to the Ovation RNA-Seq system and generate sufficient dsDNA for Illumina sequencing. Second, ex vivo amplification is laborious and could lead to bias in sample generation during the 9 days in culture. We were able to capture complete sequence coverage of the CDS directly from clinical samples therefore generating sequence data more representative of the dominant viral genome within the infected individual. Third, we utilized a recently developed de novo genome assembly algorithm, VICUNA, to assemble a full-length consensus (Yang et al., manuscript submitted). Previous studies have shown (10, 41) that de novo assembly can significantly improve the accuracy of assembly and utilization of data compared with reference-based assembly approaches. In our study, VICUNA outperformed alternate assemblers (42, 43) which can result in stalling of cDNA synthesis (44) . This uneven coverage may be due to viral secondary structure. The VICUNA assembler is able to work with this highly variable coverage to generate full-length consensus assemblies for the HIV, RSV and WNV samples.