Thursday, July 17, 2014
Do friends have similar genomes?
Looking at differences between nearly 2,000 people, recruited as part of a heart study in a small US town, they found that friends shared about 0.1% more DNA, on average, than strangers.
While small, this is the same level of similarity expected for fourth cousins.
Other scientists are sceptical about the paper, which was published inPNAS.
"I think that they're unusual findings, and that usually draws criticisms from scientists," said Prof James Fowler, one of the study's authors and a professor of both medical genetics and political science at the University of California, San Diego.
Together with Prof Nicholas Christakis from Yale University, Prof Fowler analysed nearly 500,000 single-letter markers from across the genome, using data from the Framingham Heart Study.
These results were useful because as well as providing DNA samples, participants were asked who their closest friends were. "Because the study started in a small community, many people that were named as friends, also happened to be involved in the study," Prof Fowler explained.
So he and Prof Christakis calculated a "kinship coefficient" using the genetic markers from pairs of friends and strangers, and found that it was slightly higher among friends.
"We're not really making claims about specific candidate genes here," Prof Fowler told BBC News. "We're making claims about structural characteristics across the entire genome."
Other researchers have expressed concern about different factors that could affect the results - such as ethnicity or other types of "population stratification" - which could make people both genetically similar and more likely to be friends.
Prof Evan Charney from Duke University, who has criticised earlier research by Fowler and Christakis, said this type of analysis only works if none of the subjects are related to each other at all, which is difficult to confirm.
"These studies depend upon that assumption - that you're looking at thousands of people who are not related," he told the BBC.
The authors, however, tried to allow for any stratification or family relationships within their population, in a smaller comparison of 907 pairs of friends, this time including nearly 1.5 million genetic markers.
"We excluded any one of them that had any relatedness whatsoever," explained Prof Fowler. "We didn't want anyone to think that this was just being driven by people who were accidentally friends with their fourth cousins and didn't tell us." bbc