“With the flu, every year is a rapid arms race. Countless numbers of viruses are pushed to mutate and spread over and over again, transforming until they can finally evade the human antibodies that have built up to fight off their predecessors. This shifting landscape poses a particular problem for the people in charge of reformulating the annual flu vaccine. How do you prepare for the onslaught of a virus that doesn’t exist yet?
Two computational biologists have just unveiled the first computer model that forecasts the yearly changes in the worldwide populations of flu viruses. As they report today in Nature, their model will have immediate impact in the development of flu vaccines—and proves that in some cases, projecting evolutionary change may not be beyond our reach.
‘We don’t actually predict new mutations in the flu virus,’ says Marta Luksza, one of the scientists at Columbia University, ‘Our model only considers the rise and fall of families of closely related viruses.’ Still, the computer model has proven that it can with 93 percent accuracy predict which families will harbor the most widespread viruses in the upcoming year.”