In 1918, Ronald Aylmer Fisher, an evolutionary biologist and pioneer of modern statistics, published a paper on the genetic causes of disease that brought together two rival factions. Geneticists promoted a paradigm in which diseases worked a lot like Mendel's pea plants, with just one or two genes responsible for each condition. Biometricians, however, advocated a continuous distribution of phenotypes. Fisher suggested that many mendelian traits could result in the continuous distribution of a disease. In doing so, he established the conceptual basis for the search for complex disease genes that continues today.

But Fisher's theories had a more immediate impact on animals and agriculture than on medicine — in people, it's much easier to study and measure mendelian diseases and traits. Even the much-heralded Human Genome Project in the 1990s didn't help as much as expected. The two methods traditionally used to hunt down disease genes are linkage analysis, which uses large family trees to work out which genes are shared by affected individuals, and the candidate-gene approach, which uses physiological clues to narrow down potential culprits. But when it comes to complex conditions such as heart disease or diabetes, in which multiple environmental and genetic factors combine, neither method is very powerful. Scientists have identified just a handful of disease genes, along with lots of weak, unconfirmed hits.

Now, after a shaky start, hopes are high that a more ambitious breed of genetics study can finally crack the problem. Modern gene-chip technology combined with recently published maps of human genetic variants — particularly the 'HapMap' that groups together related variants called single nucleotide polymorphisms — now enables the entire genomes of thousands of people to be scanned. Many population geneticists and disease researchers think that such genome-wide association (GWA) studies will identify genes that confer even a small extra risk of disease.

The idea of hunting across the whole genome for links to disease is beginning to pay off. Credit: P. MENZEL/SPL

Many people didn't know how much association studies would deliver.

It has taken time for big GWA studies to be completed. “Many people didn't know how much association studies would deliver,” says Peter Donnelly, a lead investigator of the Wellcome Trust Case Control Consortium, which began collecting samples for GWA studies in 2005.

Yet new results, including a study on type 2 diabetes published this week (R. Sladek et al. Nature doi:10.1038/nature05616; 2007) suggest that the GWA approach will bear fruit, and lots of it. The group, led by Constantin Polychronakos of McGill University in Quebec, Canada, studied some 393,000 single nucleotide polymorphisms in the genomes of around 700 patients with type 2 diabetes, and 600 controls; the findings were then confirmed in another 14,000 people. The researchers identified four genomic regions that confer a significant risk to developing the disease. Along with the previously identified TCF7L2 gene, these regions together might account for 70% of the genetic risk for the disease.

Of the four new genes, Polychronakos says that the best hit is SLC30A8, a zinc transporter, which is important because zinc assists with the packaging and secretion of insulin. The importance of the other hits, which have roles in pancreas development and insulin degradation, are less clear, he says.

In general, Polychronakos believes the most likely result of GWA findings will be diagnostic tests that predict who is at high risk of disease. He also envisions using genotypes to determine who would respond to which drugs, in the much-anticipated era of personalized medicine. Drug leads based on gene finds are less likely, Polychronakos thinks. Often, the disease genes uncovered are transcription factors, which he says make poor drug targets. But he does suggest that in the case of the diabetes study the zinc transporter could make a future drug target, or zinc could be used in treatment.

Geneticist David Altshuler of the Broad Institute in Cambridge, Massachusetts, is more excited about prospects for new therapies. He cites the case of cholesterol, in which a study of heart disease uncovered gene mutations related to cholesterol, allowing researchers to develop a group of drugs called statins.

Either way, the diabetes paper promises to be the first of several big finds. Donnelly says that in the next six months or so, the Wellcome Trust Case Control Consortium plans to publish genes associated with seven complex diseases, including coronary heart disease and rheumatoid arthritis. The formation of large collaborations focused on particular diseases — such as the FUSION study for type 2 diabetes — should help too, allowing researchers to share massive sets of genotype samples.

That's not to say that there aren't challenges ahead. If a gene is particularly rare, or if a disease involves dozens of genes that each have a small effect, then even sample sizes of several thousand might not pick up the signal. Donnelly, though, is more optimistic about the promise of the technique than he was three years ago. “The way I think about it is that some diseases will need much larger studies for us to be convinced of an effect,” he says.

Diseases that have a wide variety of symptoms and physiological characteristics, such as schizophrenia, may be more difficult to address. “My advice: find as homogeneous a phenotype as possible,” says Polychronakos. For example, he and his colleagues excluded obese people from their study so that they could focus on diabetes genes that confer risk independently of obesity.

Still, based on the number of papers coming up in 2007, Altshuler expects a major jump in the number of solid leads for disease genes, something neither linkage analysis nor the candidate-gene approach could match. Modern biology may finally have begun to bring technological and scientific rigour to Fisher's decades-old insights.