Evolution by numbers

Everyone on earth is human, yet all are recognisably different. The International HapMap Project aims to discover the genetic basis of human variability and its links to health and disease. In 2007 the participating scientists, from more than 70 institutions worldwide, published their latest results, based on genotyping of 270 individuals from European, African and Asian populations.

With its concentration of expertise in both genetics and statistics, Oxford is a leading centre for the analysis of these data. ‘Most conventional statistics are based on things that don’t scale’, says Professor of Statistics Gil McVean. ‘It’s a huge challenge to turn billions of data points into an understanding of where a disease gene is.’

HapMap looks not for single-letter differences in the genetic code but for patterns of variation that tend to occur together in particular regions of the genome. Because humans evolved relatively recently, many of these ‘haplotypes’ are common to all populations, while others have arisen through mutation in groups that at some point became geographically distinct.

The publicly available HapMap data provide an essential resource for more focused studies such as the Wellcome Trust Case Control Consortium, which is looking for patterns associated with common diseases including coronary heart disease, TB and diabetes; and the 1000 Genomes Project, launched in January 2008, which will examine human variation at an unprecedented level of detail.

‘We make models of the evolutionary process that has led up to what you see in the world today’, says Professor McVean. ‘If we can model these patterns accurately, it’s a very powerful tool in the analysis of the data.’ Such models, he adds, have applications much wider than simply finding links to disease or responsiveness to therapeutic drugs. ‘We’re interested in which parts of the genome are important in making us human.’