Prediction tool could help control bird flu
A $3 million study aims to develop computer models to forecast and control avian influenza outbreaks.
A tool that can understand and forecast how the bird flu virus evolves and the risk associated with new virus strains is being developed by an international research consortium.
The tool will be the first to predict factors that are currently difficult to forecast, such as whether new virus strains will cause serious harm.
Outputs could benefit vaccine development and control strategies to prevent future outbreaks.
The $3 million study, led by the Roslin Institute, is funded by the US National Science Foundation, the UK Biotechnology and Biological Sciences Research Council, and the National Natural Science Foundation of China. It is a collaboration with the US Department of Agriculture’s National Poultry Research Center, the University of Georgia and the Chinese Academy of Sciences’ Institute of Microbiology.
Bird flu can cause massive economic losses in global food production, as it infects pigs as well as poultry. The virus can also be fatal in humans.
Producing vaccines is difficult because of the variety and changeability of flu strains.
The prediction tool will combine three computer models and will use more than one million data entries of the virus’ genetic makeup, derived from global surveillance programmes.
One model will identify key changes in the virus genome that influence its behaviour, another will predict how the virus evolves within a bird as it adapts to different immune conditions, and the third will integrate outputs from the first two with additional geographic and other data.
The models will be tested for accuracy in laboratory experiments.
Development of the models will be informed by a series of workshops to gather input on viral evolution risk and predictions from stakeholders such as the World Organisation for Animal Health, the World Health Organization, and poultry holders, to ensure useful outputs.
Bird flu is a health and economic burden worldwide and tackling the virus is a big challenge. The computer models that will be developed in this study could be used to inform vaccination and other control strategies and prevent future outbreaks.
Advances in computational approaches, combined with the large amounts of sequencing data available, mean that it is possible to develop modelling tools with genuine predictive power of the evolution and spread of bird flu.
** The Roslin Institute is part of the University of Edinburgh’s Royal (Dick) School of Veterinary Studies. **