Friends often look alike. The tendency of people to forge friendships with people of a similar appearance has been noted since the time of Plato. But now there is research suggesting that, to a striking degree, we tend to pick friends who are genetically similar to us in ways that go beyond superficial features.

For example, you and your friends are likely to share certain genes associated with the sense of smell.

Our friends are as similar to us genetically as you’d expect fourth cousins to be, according to the study published Monday in the Proceedings of the National Academy of Sciences. This means that the number of genetic markers shared by two friends is akin to what would be expected if they had the same great-great-great-grandparents.

“Your friends don’t just resemble you superficially, they resemble you genetically,” said Nicholas A. Christakis, a physician and social scientist at Yale University and a co-author of the study.

The resemblance is slight, just about 1 percent of the genetic markers, but that has huge implications for evolutionary theory, said James Fowler, a professor of medical genetics and political science at the University of California at San Diego.

Lead authors, Nicholas Christakis and James Fowler, say you and your friends probably like the same scents. (Paul Schnaittacher)

“We can do better than chance at predicting if two people are going to be friends if all we have is their genetic data,” Fowler said.

This is a data-driven study that covers hundreds of friendship pairs and stranger pairs, plus hundreds of thousands of genetic markers. There’s no single “friendship” gene driving people together. There’s no way to say that a person befriended someone else because of any one genetic trait.

The research suggests that genetic factors are like a subtle breeze in the background, strong enough to be measured statistically in a big data set even if people in their day-to- day lives aren’t consciously aware of it.

Fowler acknowledges that there are limitations to the study. The scientists used data from a multi-decade medical research study of 1,932 people in Framingham, Mass., who have been participating in a heart-disease research project that dates to 1948. Almost everyone in the study group is white, and many are of Italian ancestry. The ethnic homogeneity was useful for this kind of research because it gave the scientists the chance to look for similarities among friends that go beyond something as obvious as ethnicity.

“While we’ve found that this is true for this one well-studied group of people, we don’t know if the results can be generalized to other ethnic groups,” Fowler said. “My expectation is that it will, but we don’t know.”

The scientists looked at 1,367 friendship pairs. In that group, they looked at 466,608 genetic markers and variants of those markers. They found that friends were more likely than strangers to share many of those genetic variants.

Another wrinkle: You tend to pick friends whose immune systems are strikingly dissimilar.

This might seem to contradict the initial hypothesis (that “genetic-likes”attract), but reinforces the broader thesis that there could be a subtle biological influence on friendship preference. The preference for people with markedly different immunity may have survival advantages. If you are immune to pathogen “X,” and your friend is immune to pathogen “Y,” neither of you can catch either the “X” or “Y” disease from the other.

How, exactly, do we sniff out these biologically congenial people and make friends of them? That’s not clear. We don’t wander around with gene-sequencing equipment. There may be many factors at work, some of them trivial. People with certain body types or hair color may feel more comfortable with one another.

And friendships could develop around certain locations. People who love the smell of coffee may hang out in coffee shops and wind up becoming friends with other coffee aficionados, and it’s not surprising if some of these friend-pairs have similar olfactory genes. Swimmers might make friends at the beach with other aquanauts.

This is not a settled science. Research on genetic factors in friendships is still in a preliminary stage. But if the reasoning of Christakis and Fowler is correct, friendship, and hyper-social behavior more generally, is a significant factor in the recent evolution of the human species, which they describe as having accelerated in the past 30,000 years.

We think of evolution as a process driven by natural selection among individual organisms. But natural selection pivots on the fitness, or reproductive success, of specific genes. It’s long been understood that this requires us to look at kinship groups when thinking about the reproductive success of those genes.

This new theory says that it doesn’t look broadly enough at how evolution works. Our friends are also in the Darwinian pool with us.

Your evolutionary fitness “depends not only on your own genotype, but also on the genotype of your friends,” Christakis said.

“Social networks are an important engine for human evolution,” Fowler said. “Our friends are sort of like family members. They’re functional kin.”

Robert Seyfarth, a professor of psychology at the University of Pennsylvania, who was not involved in the new research, said of the study, “This is a very interesting, provocative answer to the question of why is it that humans are so hyper-social in their interactions. Why are they so friendly to strangers? Most animals don’t encounter strangers at all.”