Humans Choose Friends With Similar DNA

KentuckyFC writes "The study of social networks has long shown that people tend to pick friends who are similar to them — birds of a feather stick together (PDF). Now a study of the genomes of almost 2000 Americans has found that those who are friends also share remarkable genetic similarities. 'Pairs of friends are, on average, as genetically similar to one another as fourth cousins,' the study concludes. By contrast, strangers share few genetic similarities. The result seems to confirm a 30-year-old theory that a person's genes causes them to seek out circumstances that are compatible with their phenotype. If that's the case, then people with similar genes should end up in similar environments and so be more likely to become friends."

Re:Fourth cousins?

[alternativity]

Not really. The idea of six degrees of separation is based on acquaintance and not on genetic relations. E.g. think of China under the one child policy, if it continues, then each passing generation will lose a subsequent degree of cousins. The first cousins would disappear first, then second, third, fourth and so on, but six degrees of separation will still apply to them.

AC, party of one

[Valdrax]

I've lived around the world and have a variety of friends.

Well neat. You alone must be a representative sample set of humanity that dwarfs this mere 1932 person sample group for statistical relevance. Thank goodness you know more about statistics than people doing a population study!

I'm assuming these 2000 individuals for the study were in environments that just happened to have similar genes around them.

Actually, if you read the study (OMG! I must be new here!), then you'll see that they address this point:

"There are (at least) four possible reasons that friends may exhibit homophily in their genotypes. First, correlation in genotypes may be a trivial by-product of the tendency of people to make friends with geographically proximate or ethnoracially similar individuals who also tend to share the same ancestry. Thus, it is important to use strict controls for population stratification in tests of genetic correlation (below, we rely on the widely used principal components method to control for ancestry). [...] Third, people may actively choose particular environments, and, in those environments, they may be more likely to encounter people with similar phenotypes influenced by specific genotypes. If people then choose friends from within these environments (even at random), it would tend to generate correlated genotypes."
[...]
"To eliminate the possibility that the results are influenced by people tending to make friends with distant relatives, we use only the 907 friend pairs where kinship â 0 (recall that kinship can be less than zero when unrelated individuals tend to have negatively correlated genotypes)." (ed: They do the same with 907 stranger pairs.)

In the end, they find that people prefer friends that share genes for the same sense of smell and the same linoleic acid metabolism. We also strongly prefer people with different genotypes for immune system function. While there are hundreds more homophilic and heterophilic gene correlations, those were the three that were most over-represented between people. There are many other genes that friends share, but most of those vary from pair to pair of people and are either idiosyncratic preferences or possibly just coincidences. Those three genes are not.

It's pretty well known that human select their lovers in part by smell. It's interesting that we pick our friends too that way. The paper gives a few good theories for why that's true (based on older research) and also does so for the immune system. The linoleic acid thing seems to have baffled them a bit, though they make a stab at explaining it as possibly being related to food safety and keeping the community on the same page about what's good to eat.