Understanding the emergence of bird flu: epidemiological and phylogenetic analysis of the global spread of H5Nx lineages of avian influenza

Avian influenza viruses have been a major global public health concern ever since the first emergence of a H5N1 ‘bird flu’ strain in the 1990s that not only had a devastating impact on poultry populations but also caused an often fatal disease in humans. Since then, several other strains of avian influenza have been reported in humans. Though some of these can also cause severe disease none of them have yet proved to be transmissible within human populations.

Following an outbreaks in eastern Asia in 2014, a subtype H5N8 virus (known as clade 2.3.4.4), rapidly spread worldwide in 2014-5 [1], and then again in 2016-7, with migratory birds playing a major role. Additionally, haemagglutinin of clade 2.3.4.4 virus is remarkably promiscuous compared to other H5 segments and has successfully reassorted with at least six different neuraminidase subtypes (collectively referred to as H5Nx), including H5N6 which has recently been found in human cases. This matters because the combination of segments in an influenza virus, especially haemagglutinin (H) and neuraminidase (N), determine many of its phenotypic characteristics, including its host range and transmissibility. An important but as yet unanswered question is to explain the unusual promiscuity of clade 2.3.4.4 H5 haemagglutinin and to determine whether it is linked to its pathogenicity characteristics (these viruses cause severe illness in some hosts but mild infections in others).

In this project, the student will test two hypotheses concerning H5Nx viruses. The first is whether these viruses have an especially highly ‘evolvable’ cell receptor specificity, which would be indicated by sequence variation along lineages of the haemagglutinin gene itself. The second is whether these viruses interact with the host cytokine pathways in ways that differ from closely related viruses, which could be indicated by sequence variation in segment 8 of the influenza genome.

The project has access to several thousand influenza genome sequences held on public databases, as well as early access to recently obtained sequences through the GISAID platform and the Global Consortium on H5N8 and related influenza viruses.

Data analysis will build on previous work and will involve a combination of genome sequence analysis, evolutionary models, computational biology and machine learning methods, likelihood based tests and approximate Bayesian computation. The student will receive one-to-one training in the use of state-of-the-art software platforms for implementing these powerful techniques. In addition, broader training will be provided, as required, in bioinformatics, computational biology, data science and epidemiology, plus access to considerable expertise in molecular virology.

Good computational and statistical skills are required, as well as an interest in applying these skills to biological problems of global public health significance. The successful applicant will not necessarily be familiar with the analytical approaches needed, but should be able to demonstrate an aptitude and enthusiasm for learning new skills and applying them to the rapidly expanding global resource of genome sequence data. A background in biology, genetics or epidemiology with good quantitative skills would be suitable, as would a background in computation or physics with an interest in biomedicine.