The thorny question of race is set to take centre stage at the annual meeting of the American Society of Human Genetics in Toronto, Canada, this week. Leading geneticists will present perspectives on how their discipline should deal with the fraught topic.

The geneticists' views will be published in a supplement to next month's Nature Genetics, and were to be the subject of a press conference that was due to take place in Toronto on 27 October. Their views took shape at what many participants characterize as a landmark meeting in 2003 at Howard University, a historically black university in Washington DC.

The scientists' overall message is that “the traditional categories of race and ethnicity are to some extent inhibiting researchers from identifying the real environmental and genetic causes of disease,” says Myles Axton, editor of Nature Genetics. “The genetic differences that are of importance in disease are not necessarily distributed according to race and ethnicity.” But the authors of the supplement, whose production was sponsored by the US Department of Energy, are sharply divided on how quickly they can afford to discard ‘race’ as a valuable category in their work.

Francis Collins, director of the US National Human Genome Research Institute, who contributed a commentary article, says that the connections between race, genetics and disease are ripe for serious scientific study.

Stir it up: does race have a genetic basis that could aid biomedical research? Credit: J. WIEDEL/ALAMY

Unlike some of his fellow authors, Collins does not think that the concept of race should be ditched. “Throwing it away might cause us to lose some of our best clues of the causes of disease, be they genetic or environmental,” he says. However, he adds, race is “a lousy proxy” for specific risk factors and researchers should dispense with it as soon as they identify pertinent factors. “It's a lot more powerful to test for sickle-cell carrier status than to look at the colour of somebody's skin,” he says.

Human beings differ from each other at only about 0.1% of the genome's 3 billion bases, but it is not known whether or how those differences create disparities in disease prevalence, severity and drug response between different races. Some argue that studying this question will provide vital information for biomedical progress. Others say that this approach relies on unvalidated assumptions that race has a biological base, and plays into the hands of racists.

“Genetic variation research does not support the existence of human races. That is our overriding consensus,” says contributor Charles Rotimi, acting director of Howard's human genome centre. “The only way we can get to the meaning of human genetic variation is by removing racial classification in biomedical research.”

“We need to make changes,” agrees Charmaine Royal, a geneticist at Howard and the prime mover behind the 2003 meeting. Study designs, she suggests, should reflect subjects' geographical location and ancestry, for example, rather than their race.

But Neil Risch, a geneticist at Stanford University, California, strongly disagrees. “It's extremely important to examine disease rates in different racial groups,” he says. “There is no way to address inequities in health otherwise.”

Joanna Mountain, a Stanford anthropological geneticist, says that nearly all researchers agree that, ultimately, race and ethnicity need to be discarded as categories in biomedical research. “But in the short term, it may be far more efficient to use them than to ignore them. Race and ethnicity are explanatory, even if it's unclear what they are surrogates for.”