Genes? It's complicated
The advance heralded a decade ago in mapping human DNA is yet to lead to the answers we craved

Ten years ago the $10bn Human Genome Project announced it had completed the first draft of the blueprint for human life. It was hailed as a huge scientific advance, comparable to putting a man on the moon. President Bill Clinton declared: "We'll go from knowing almost nothing about how our genes work to enlisting genes in the struggle to prevent and cure illness. This will be the scientific breakthrough of the century, perhaps of all time."

The project at last laid bare the entire human genetic code – 22,000 or so genes (the precise number is still uncertain) – that make us into the people we are. Several decades of research into the cause of diseases before the project had firmly identified genes as a significant cause of many important diseases.

The first haul of genetic diseases was of those fairly rare but devastating inherited diseases, such as cystic fibrosis and haemophilia, that are caused by single genes. Most of the genes responsible for those had been fished out of the genome long before the sequencing project hauled in its net. But the project was expected to find genes for various far more common conditions, such as cancer, diabetes, heart disease, autism, depression and schizophrenia, because most of these conditions tend to run in families. Studies of families in which these diseases were common, particularly of twins, had established a level of heritability for each condition, and the levels were high. Autism comes out at a whopping 90%, indicating that most autism is caused by faulty genes (and certainly not by faulty vaccines). The heritability of schizophrenia was about 80% whereas conditions such as heart disease, diabetes and cancer came in anywhere between 30% and 70%.

And it wasn't just diseases that were caused by genes. Many behavioural studies indicated that intelligence, personality, sexual orientation and even voting preference seemed to be highly heritable. If genes were so powerful, it should be straightforward to identify the culprits in the genome.

But a decade later these expectations have not been fulfilled. The project that promised so much has, so far, delivered very little. Very few genes have been found that account for more than 1% of the risk of any of those common diseases. And even the most significant intelligence gene yet found is responsible for variation in individual intelligence equivalent to less than one IQ point. The scientists who went in search of whoppers netted only a host of minnows. Where are the missing genes?

Like most things in life, it turns out that genes are more complex than we thought. Those genes responsible for single-gene defects such as cystic fibrosis and haemophilia are the low-hanging fruit. Common diseases, and such attributes as intelligence, are not caused by single genes or even handfuls of genes, but probably by networks of hundreds or even thousands of genes.

To understand these networks, we need to look, not at the branches, but at the roots of the genetic tree. Genes form tangles of interactions with each other such that the effect of chopping one or another is unpredictable and depends on the connectivity of the whole network. Finding a gene responsible for a disease is mostly like finding a root responsible for maintaining a tree.

The task of unravelling the roots of biology is the new science of system biology, in which biologists work with mathematicians and computer scientists to build models of complex networks. This is where the causes of heart disease, diabetes and autism are now being sought. To paraphrase Winston Churchill, the genome project was not the end. It was not even the beginning of the end. But it was, perhaps, the end of the beginning in the search for our genes.