Integrating patient and whole genome sequencing data to provide insights into the epidemiology of seasonal influenza A(H3N2) viruses

Goldstein EJ, Harvey WT, Wilkie GS, Shepherd SJ, MacLean AR, Murcia PR and Gunson RN (2017) Microbial Genomics DOI 10.1099/mgen.0.000137 PDF

Abstract

Genetic surveillance of seasonal influenza is largely focused on sequencing of the haemagglutinin gene. Consequently, our understanding of the contribution of the remaining seven gene segments to the evolution and epidemiological dynamics of seasonal influenza is relatively limited. The increased availability of next generation sequencing technologies allows rapid and economic whole genome sequencing (WGS) of influenza virus. Here, 150 influenza A(H3N2) positive clinical specimens with linked epidemiological data, from the 2014/15 season in Scotland, were sequenced directly using both Sanger sequencing of the HA1 region and WGS using the Illumina MiSeq platform. Sequences generated by the two methods were highly correlated and WGS provided on average >90% whole genome coverage. As reported in other European countries during 2014/15, all strains belonged to genetic group 3C, with subgroup 3C.2a predominating. Multiple inter-subgroup reassortants were identified, including three 3C.3 viruses descended from a single reassortment event, which had persisted in the population. Cases of severe acute respiratory illness were significantly clustered on phylogenies of multiple gene segments indicating potential genetic factors warranting further investigation. Severe cases were also more likely to be associated with reassortant viruses and to occur later in the season. These results suggest that WGS provides an opportunity to develop our understanding of the relationship between the influenza genome and disease severity and the epidemiological consequences of within-subtype reassortment. Therefore, increased levels of WGS, linked to clinical and epidemiological data, could improve influenza surveillance.

Blog post: Analysis of flu full genomes with linked patient data

BioRxiv preprint