Identification of viral transcripts in RNA-seq datasets from bees, ants, wasps and mites

Abstract

Many honey bee colonies suffer large losses due to colony collapse disorder. This phenomenon, which has dramatically increased in frequency since 2006, has led to widespread efforts in sequencing honey bee pathogens, including RNA viruses such as deformed wing virus. However, honey bees coexist with a number of other arthropods, whose viruses are less thoroughly characterised. Many viruses currently classified as honey bee pathogens may therefore have a much wider host range. In particular, ants, which like bees are members of the Hymenoptera order, often coexist with bees and the two groups have previously been shown to exchange viruses. Parasitism by Varroamites, known to act as effective vectors for a number of RNA viruses, is also almost ubiquitous amongst honey bees, but little is known about viruses endemic to mites. We have previously demonstrated that it is possible to detect and characterise viral RNA in publicly available RNA-seq datasets. There are over 3000 such datasets for diverse Hymenoptera and mite species. We have developed a computational pipeline to identify viral transcripts in these datasets. This pipeline performs quality control, removes low complexity reads and reads generated from host RNA and various known contaminants, assembles the remaining reads into transcripts and detects the presence of regions with homology to known RNA viruses. Viral fragments identified with this pipeline will be examined phylogenetically to identify novel pathogens, clarify host range and specificity, and characterise transmission patterns.