Generalizing Herpesvirus Genome Assembly: A Novel Bioinformatics Approach

Herpesviruses, such as herpes simplex virus type 1 (HSV-1) and Epstein-Barr virus (EBV), are highly prevalent DNA viruses that can cause recurring, lifelong infections and are associated with various cancers. Due to its prevalence in humans and other species, it is vital to study these viruses to gain a better understanding of how they infect and persist in their hosts. Herpesviruses were originally regarded as viruses with low genomic diversity, but as more herpesviruses are sequenced with a higher accuracy, that paradigm has started to shift as new research shows that they can evolve faster than previously assumed.

However, these viruses are notoriously difficult to assemble due to their large genome size (125-240 kb in length), extremely high GC content, and numerous terminal and internal repeat regions. With available short-read only approaches, resulting assemblies are often fragmented; on the other hand, genomes assembled using long-read only approaches are more complete, however, they often fail to capture the inverted terminal repeat (ITR) regions. Publicly available herpesvirus genomes tend to use a more manual approach that involves laboriously curating each contig. Due to manual intervention, these assemblies are more complete; however, for high-throughput labs, such methods cannot be easily employed. 

Here, we present our assembly pipeline which leverages long-read Oxford Nanopore (ONT) data and includes herpesvirus specific pre-processing steps to generate complete herpesvirus genomes. Briefly, long reads are first trimmed according to ATCC’s quality standards, and the resulting reads are then binned based on taxonomy use Kraken2. Any reads that classify as “virus” move forward to assembly using the Flye assembler. Uniquely, this pipeline does not employ de-hosting of the reads as this process can remove essential herpesvirus reads as the human host genome contains sequences homologous to herpesviruses. Genomes assembled using this pipeline were more complete and resolved the ITR regions. Through this method, we were also able to assemble more complete genomes of other DNA viruses, such as adenoviruses.