Landmark sequencing project makes predicting flu into realm of the possible

TORONTO (CP) - A landmark project to chart the genetic blueprints of thousands of human influenza viruses - essentially taking a genetic census of the flu - has uncovered why the 2003 flu vaccine didn't protect very well against the A/Fujian strain that dominated that winter.

U.S. researchers mining data from the sequences found that a minor cousin and the dominant flu strain swapped genetic material in the spring of 2003, after the World Health Organization's panel of experts chose the dominant strain for the following winter's vaccine.

Since it was wearing its cousin's winter coat - a surface gene called hemagglutinin, the H in a flu strain's name - the dominant strain managed to slip passed the protective antibodies created by the vaccine.

Solving the mystery of the inadequacy of the 2003-04 flu vaccine is just the first in what experts believe will be an avalanche of discoveries to come from the vast array of fully sequenced flu viruses, sequences that have been deposited in public access databases for all the world's scientists to study.

"We've seen that as genomes come out, they lead to a revolution in the biology of many organisms, with many, many papers coming out subsequently based on genome sequence data," says Steven Salzberg, senior author on the paper announcing the A/Fujian discovery.

The paper, which was based on analysis of 209 total sequences collected in New York state over the past several years, was published Thursday in the journal Nature.

In the time since the paper was submitted, almost double that number of sequences have been completed and logged in databanks.

"We're hoping that by just cranking out huge quantities of flu sequence that many people who are somewhat restricted in what they can do right will suddenly look at this incredible resource and start doing some experiments and doing more follow on studies of this data," he said.

In fact Salzberg confidently predicts the project will eventually eclipse another influenza milestone published in the same issue of Nature, the sequencing of the final three genes of the virus that caused the worst infectious disease event in known history, the 1918 pandemic known as the Spanish flu.

"The scientific community and the medical community will learn a lot more from this project and all these genomes than from a couple more segments of the 1918 virus," said Salzberg, director of the Center for Bioinformatics and Computational Biology at the University of Maryland.

The project, which is funded by the U.S. National Institute of Allergy and Infectious Diseases, will eventually put thousands of flu virus sequences at the disposal of the research community.

Promise of access to this wealth of flu genetic information is like Christmas coming early, said Joshua Plotkin, a computational evolutionary biologist at Harvard University.

Plotkin is researching ways to better predict how influenza viruses mutate - a goal, if achieved, that would help flu vaccine designers avoid vaccine failures like the one in 2003.

"The really obvious sort of upside of that is if you do have a really good feeling for those patterns and why they arise, then you actually do have a rational basis for choosing vaccine strains - as opposed to best guesses," he said.

Until this project began, getting sequence data for flu viruses was a difficult task. Public databases contained only a few hundred sequences, and most of those covered only one or two genes from each virus, not the full eight.

The limits on the available data constrain the ability of scientists to actually see what is happening with influenza, says Earl Brown, a virologist from the University of Ottawa who specializes in evolution of influenza viruses.

"You want to see as much of the picture as possible. And a lot of it's hidden just because people aren't sequencing or looking at it," Brown said.

Recently, the U.S. Centers for Disease Control in Atlanta came under fire for not sharing sequencing information from its large bank of viruses. A news article in Nature quoted a number of anonymous researchers complaining the CDC's slowness to share was impeding research.

Salzberg said genomic sequencers - of which he is one - came to the painful realization about five years ago that it was no longer acceptable to sit on sequence data until a team had time to analyze it and write the ensuing scientific manuscripts.

Researchers simply could not keep up with the flow of new sequence data, he said.

"So if we sit on it, which is the traditional way of doing science, until our papers are out - or longer - then the science goes slower."

He believes it's time for the flu community to come to the same conclusion. But some have been slow to buy in to the idea.

"The real challenge is getting samples, getting people to share their samples. Because the deal is when you give us your sample collection, we're going to release all the data.

"Flu people kind of like to have their samples and keep them to themselves. This is not just a U.S. thing. This is worldwide," he said.