[YY-EEID US-UK XXXX] Predictive phylogenetics for evolutionary and transmission dynamics of newly emerging avian influenza viruses

Influenza virus is a global problem, causing widespread harm to human health and the food production system because it also infects chickens and pigs. Vaccination is difficult because of the variety and changeability of flu strains found in nature - primarily in wild birds, where often they cause little harm. However, when these strains of virus spill over into domestic poultry or humans, they can cause massive economic losses and fatal disease respectively. In the last twenty years, this has been graphically illustrated by the H5N1 and H7N9 outbreaks. Global surveillance programmes track the virus' movement and as part of this, characterise the sequence of the viral genome. Some aspects of virus behaviour can be accurately predicted from these sequences. However, many other important aspects of virus biology, such as whether it will travel across continents, which species it will infect and whether it will cause serious harm, are much harder to forecast. Our premise is that the volume of sequencing data now available, along with recent advances in computational methods of using such data, will make it possible for the first time to generate virtual models of how the virus will evolve under specific circumstances and how these viral variants will behave. Such models have the potential to produce risk estimates of new strains as they arise that can be used to inform policy and direct strategies to head off impending threats. To achieve this goal, we have brought together a team of international experts with interdisciplinary expertise in mathematical modelling, influenza surveillance and biology, and the infectious disease-public and animal health interface. Importantly, this includes colleagues from China, the likely epicentre of the virus. Together, we will create the computer models that can understand and forecast virus evolution; models that will be made accurate and then tested through a series of focussed laboratory experiments designed to produce the needed data, and whose types of output will be tailored to the needs of end users through a series of workshops that include the primary stake holders so they can inform the scientists on what information they need.