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.

We assembled Illumina data using the VICUNA assembly program (Yang et al., manuscript submitted). In short, VICUNA used the Ovation RNA-Seq primer sequence (inferred from several input datasets) for read trimming. Low complexity reads were discarded. Since some of the HIV, RSV and WNV datasets contained a large percentage of reads derived from the host, we invoked the optional target genome filter component of VICUNA. Based on sequence similarity, the remaining reads were clustered to form contigs, which were further validated such that each read in the contig had sufficient similarity compared with the consensus. Guided by paired reads, these contigs were compared using an efficient alignment algorithm. Any two contigs that shared a significant suffix-prefix overlap were merged. The assembly process terminated when no further merging could be applied. In several cases, a viral reference genome was used to further extend the assembly. Briefly, raw contigs of >350 bases obtained by the assembler were aligned to a reference using MUSCLE (version 3.8) (38) . Each contig was broken down in segments when there was a deletion of 30 or more nucleotides compared with the reference. The contigs were added to form the final assembly in decreasing order of length until either the whole assembly was covered or no contigs remained. Overlapping segments were merged together by concatenating the nucleotide sequence of each segment at the central position of overlap.