Predicting the flu of the future

This will enable bodies such as the World Health Organisation to stock up on suitable vaccines and reduce the global impact of any epidemics.

The new approach came from constructing and examining a decade-long family tree of the influenza A viruses.

Testing the approach using past data, the scientists correctly predicted the dominant strain of flu in nine years out of 11. The chances of doing this by random guessing are tiny, just one in two billion.

"Scientists haven't known how to predict which strain of influenza virus is going to be the progenitor of the strains that will cause future epidemics," said Walter Fitch, at the University of California Irvine, who lead the research team. "This is the first time that an evolutionary study has been used to identify which strains are the fittest."

A periodic table of biology

David Hillis, at the University of Texas, commenting on the research published in the journal Science, said: "Just as the periodic table of elements allows chemists to make predictions about reactions, so a [family tree] allows biologists to make predictions about behaviour or any other biological attribute."

He added: "The team show that analysis of the evolution of human influenza virus A can be used to make predictions about the evolutionary course of future strains."

Influenza viruses come in three types - A, B and C. Influenza C has very little public health impact, but influenza A and B viruses cause the annual epidemics of influenza seen in many parts of the world. The research team is now extending its work to include all three groups of circulating influenza viruses.

Spindly tree

The new method of prediction works because of the flu virus's unusual family tree. While most organisms evolve through time to create a tree with many forked branches, the flu virus tree shows that in each generation just one main strain survives. This leaves the tree looking more like a stepped staircase.

This unusual pattern of lineage occurs because, to be successful, the flu viruses have to evade the human immune system which will attack and destroy anything it recognises as foreign.

The viruses disguise themselves using spikes on their surface made of a protein called hemagglutinin. But once the body has seen a particular spike, it will remember it and attack it if the virus infects the body again.

Therefore, those viruses which most successfully mutate the hemagglutinin spike will do best the next year. Those which do not will die out.

Change or perish

Professor Bush and his team identified 18 key regions of the hemagglutinin gene. The viruses with the most changes in these areas were the most likely to be the start of a dynasty. So by focusing on these areas, predictions can be made.

Professor Hillis believes the success of this work, in using an evolutionary approach to tackle future health problems, holds a lesson for those in the US who have imposed bans on the teaching of evolution in schools.

"School boards and science educators need to understand this simple fact: if students don't learn about evolution, they can't possibly understand modern biology or medicine."