COLD SPRING HARBOR, NY, May 30 - GeneSweep, a friendly wager in which participants bet on the number of genes on the human genome, is officially over, and there are three winners -- although genome scientists have not definitively agreed on a number of genes in the human genome.
Ewan Birney, the European Bioinformatics Institute scientist who led development of the Ensembl gene browser, opened the books on this highbrow bet at the 2000 Cold Spring Harbor Genome Sequencing and Biology conference, and tonight, almost exactly three years later, he pronounced the books closed during his talk -- as he promised he would when he first introduced the contest. Birney said that there were 24,500 genes in the latest Ensembl build, human build 33. But he stopped short of saying that this number represents any final tally on the number of human genes. "We are confident we have about 21,000 genes in the human genome," he said. "We are just not that confident about whether there are other genes."
In GeneSweep, no participants bet that there were 24,500 or fewer genes. But Birney decided to abide by his rules -- not those of The Price is Right or blackjack (which throw out any overbids) -- and declare three winners, those who were closest to 24,500 in each of the past three years of the meeting that the wager was open. These winners included Paul Dear, of the UK's Medical Research Council, who guessed in 2000 that there would be 27,462 genes; Lee Rowen, of the Institute for Systems Biology, who guessed in 2001 that there would be 25,947 genes, and Olivier Jaillon, of Genoscope in France, who guessed in 2002 that there would be 26,500 genes. Dear gets half of the pot, of which Birney has not yet figured out the total amount, and the other half of the pot is divided between Rowen and Jaillon.
"I tend to be very conservative" said Rowen, in explaining her winning strategy. She said she guessed low because she knew of 15,000 genes, and was wondering, "where are the other 15,000?" But she tempered her reaction with scientific skepticism, wondering aloud whether she would have to give back the money if it is discovered in the coming years that there are significantly more genes.
Birney, for his part, said he would not require winners to give back the money if current predictions prove to be wrong, but said that those who guessed higher would win "a moral victory" -- which is perhaps the sweetest type in science. Birney is not concerned that existing methods of gene finding have failed to locate any genes of known biological significance. "What we're worried about is the so-called dark matter," he said. "The real debate is [about] dark matter single-exon genes."