We share DNA with friends, study claims

Our choice of friends may be influenced by our genes, a controversial new study claims.

Scientists from the University of California found that groups of friends tended to carry the same version of a gene previously linked to alcoholism. Conversely, people with similar versions of a second gene, previously linked to the 'openness' personality trait, were more likely to avoid each other.

Professor James Fowler, who led the research, explained: 'That feeling that you get that you're just going to like somebody or not going to like them - a lot of times we'll have those instincts about people and we're not sure where they come from. We think that understanding the [genetic markers] that underlie friendship may help us to understand more of that process'.

Researchers examined the frequency among friends and non-friends of one genetic marker, called a SNP (single nucleotide polymorphism), in each of six genes. The same version of dopamine gene DRD2 - believed to influence alcohol habits - tended to cluster among friends. It may, however, also be the case that people who drink more tend to make their friends in pubs.

The opposite effect was observed for a second gene, CYP2A6, believed to affect personality 'openness'. Friends tended to have different versions of this marker, perhaps lending to the popular 'opposites attract' belief. Factors which may influence the results, such as gender, geographical region and common ancestry were accounted for.

However, the study, published in the Proceedings of the National Academy of Sciences, is controversial, with several geneticists questioning the robustness of the findings and the fact that only six markers were studied.

'If this was a study looking for shared genes in patients with diabetes, it would not be up to the standards of the field', said Dr David Altshuler, a geneticist at the Broad Institute in the United States. 'We set these standards after ten years of seeing so many irreproducible results in gene-association studies'.

'Because most genes have modest effects on behaviour or health, many scientists assume that thousands of SNPs - rather than six - need to be analysed before a correlation to any trait can be confidently made. Geneticists are often hard-pressed to find one SNP in a million that reproducibly correlates with a disease'.

Professor Stanley Nelson, a human geneticist at the University of California, added: 'It certainly is a provocative study - I would have loved to have seen it done with information from the rest of the genome'.

Professor Fowler refutes these criticisms, stating that genome-wide information was not available and insisted that the controls for common ancestry were extremely stringent.