Development and evaluation of a bioinformatics approach for designing molecular assays for viral detection

BACKGROUND: Viruses belonging to the Flaviviridae and Bunyaviridae families show considerable genetic diversity. However, this diversity is not necessarily taken into account when developing diagnostic assays, which are often based on the pairwise alignment of a limited number of sequences. Our objective was to develop and evaluate a bioinformatics workflow addressing two recurrent issues of molecular assay design: (i) the high intraspecies genetic diversity in viruses and (ii) the potential for cross-reactivity with close relatives. METHODOLOGY: The workflow developed herein was based on two consecutive BLASTn steps; the first was utilized to select highly conserved regions among the viral taxon of interest, and the second was employed to assess the degree of similarity of these highly-conserved regions to close relatives. Subsequently, the workflow was tested on a set of eight viral species, including various strains from the Flaviviridae and Bunyaviridae families. PRINCIPAL FINDINGS: The genetic diversity ranges from as low as 0.45% variable sites over the complete genome of the Japanese encephalitis virus to more than 16% of variable sites on segment L of the Crimean-Congo hemorrhagic fever virus. Our proposed bioinformatics workflow allowed the selection—based on computing scores—of the best target for a diagnostic molecular assay for the eight viral species investigated. CONCLUSIONS/SIGNIFICANCE: Our bioinformatics workflow allowed rapid selection of highly conserved and specific genomic fragments among the investigated viruses, while considering up to several hundred complete genomic sequences. The pertinence of this workflow will increase in parallel to the number of sequences made publicly available. We hypothesize that our workflow might be utilized to select diagnostic molecular markers for higher organisms with more complex genomes, provided the sequences are made available.