Friends are as genetically similar as fourth cousins

http://www.pnas.org/content/111/Supplement_3/10796.long

by gwern • 10 years ago

281 133 10 years ago

douche• 10 years ago

Where I grew up, it's a pretty good chance all my friends were fourth or fifth cousins. Small population + limited outbreeding = family hedges, instead of family trees.

There were kids in my grade who couldn't date in town because all the girls were their first or second cousins...

mnbcvx2• 10 years ago

second cousins

Didn't stop my great grandparents :) My mom's side is Mennonite, which is also a very tight knit community. When my grandma sends random pictures from in and around Goshen, Indiana, my wife has commented multiple times "whoa that guy in the background really looks like you!"

runholm• 10 years ago

Really not an issue with second cousins. Second cousins are genetically different enough that issues with the children are just barely more common in non-related couples.

roninb• 10 years ago

In case you were also wondering what a "fourth cousin" was, someone who shares a pair of great-great-great grandparents[0]. Wikipedia has a great set of diagrams[1] in case you're also confused about first, second, or third cousins. I figure through induction you can presume any nth and nth removed from that page (I foolishly thought 3rd was where people arbitrarily stopped keeping count....)

[0] https://www.reference.com/family/fourth-cousin-7b40e096c6d65...

[1] https://en.wikipedia.org/wiki/Cousin#First_cousins

skylan_q2• 10 years ago

Where our family is from the priests wouldn't allow marriage between 6th cousins. My grandfather was 6th cousins three times removed from my grandmother. (so she was 9 generations away from common ancestors with him) They had to get written permission from the bishop to get married.

milesskorpen• 10 years ago

Sounds like they basically penalized record keeping ...

jacobolus2• 10 years ago

The Church worked very deliberately over centuries to reduce the influence of families / clans, in some places more successfully than others. It all revolves around inheritance rights and accumulated power. (Also in some cases the Church could collect money by charging for exemptions.)

https://en.wikipedia.org/wiki/Consanguinity#Religious_and_tr...

LukeShu• 10 years ago

I used to really struggle with those types of diagrams. Like, they make sense on their own, but it's hard to reason about the patterns with them; the relationships (heh) between each diagram aren't clear.

I much prefer the square/diamond type diagram that's further down in the article. It seems like black magic at first, but if you stop and think about it for a bit, once you understand it, is crystal clear.

I guess it's about short-term or long-term ease of understanding.

ams6110• 10 years ago

Wow, takes me way back to an undergrad programming assignment to take a family tree and compute the degree and "removedness" of any two cousins.

fiatjaf8• 10 years ago

I get mad at people who think sons of first-cousins are second-cousins all the time -- and, frankly, almost everyone I know think this way.

I'm glad you linked to this Wikipedia article that proves I'm not mad into thinking there is a simple logic pattern that can be applied to cousinship.

However, this "once removed" notation is crazy. Why not second-uncle for parents of second-cousins and so on?

dietrichepp2• 10 years ago

Pardon me for my ignorance, but wouldn't sons of first cousins be second cousins?

srtjstjsj• 10 years ago

There's an app for that!

https://www.microsoft.com/en-us/store/apps/cousinosity/9nblg...

rat87• 10 years ago

That's what I use.

2nd uncle, 2nd grand aunt

Fun fact: in Russian a first cousin is a 2nd Brother/Sister, a 2nd cousin is a 3rd Brother/Sister and so on.

Throwaway23412• 10 years ago

Edit: Nvm. I am more confused than I thought.

eru3• 10 years ago

You are hitting the borders where logic and common usage meet.

I am actually surprised that descriptive grammar holds up so well to logic. (There's no a priori reason why descriptive grammar, that describes common usage of language, actually is as mostly compressible to simple rules as it is.)

moron4hire• 10 years ago

You get mad? Really? Why?

thegenius2000• 10 years ago

Worse even, the "once removed" notation has no notion of direction, so it's inherently ambiguous. I'm also pretty sure that less half of the people I hear using it know what it means (or tries to mean).

return02• 10 years ago

This got me confused. You actually mean people who confuse "second niece" (i.e. your cousin's son) with "second cousin". Is that usage common in english? The daughters of cousins are second cousins,no?

unethical_ban• 10 years ago

That's how we've always done it.

* Brother of father is uncle. Son of uncle is cousin. Son of son of uncle is second cousin.

* Brother of grandfather is great-uncle. Son of great-uncle is second cousin.

justinlardinois• 10 years ago

Isn't the term "uncle" relatively recent? I recall Hamlet referring to his father's brother as his cousin.

avodonosov4• 10 years ago

Let's calculate. I have a pair parents, 2 pairs grandparents, 4 pairs of great grandparents, 8 pairs great-great grandparents, and 16 pairs great-great-great grandparents.

How to estimate the number of fourth cousins?

logfromblammo3• 10 years ago

Through induction, you can also say that your sibling is your nilth (0th) cousin, and your identical twin is your negative-first cousin. That makes aunt or uncle a nilth cousin once removed, and a great aunt or great uncle a nilth cousin twice removed. Your parents are negative-first cousins once removed.

If this sounds ridiculous, that's because it is, as a natural consequence of the ridiculous nomenclature.

It would be far less confusing to just insert an unordered pair of numbers for both individuals' genetic distance from their common ancestor. So instead of a first cousin once removed, you would have a (2,3)-cousin. Identical twins would be (0,0), parents would be (0,1), ordinary siblings (1,1), aunts and uncles (1,2), and so on. A fourth cousin would be (5,5).

But that doesn't account for half-siblings, so it would be inadequate for some family relationships.

phyllostachys• 10 years ago

It is mentioned in some comments of comments further down but Wolfram Alpha has some great family tree generation.

http://www.wolframalpha.com/input/?i=fourth+cousin

JoeAltmaier• 10 years ago

Also, its grand-uncle (not great-uncle). You don't have a great-mother do you? Its a grandmother. Thus granduncle.

ilaksh• 10 years ago

Could it just be more common for people who are in the same social group to be in the same ethnic group, and maybe that explains most of the genetic similarity they saw, rather than necessarily proving some kind of fitness advantage or something? Probably because often your friends are from the same neighborhood.

aab0• 10 years ago

The genetic similarity here is estimated within the cohort; the Framingham cohort is by design ethnically homongeous to try to eliminate that sort of population structure confound. This way, the homophily is not picking up on the obvious stuff like ethnic groups. If they studied people from multiple ethnic groups, the results would be trivial and uninteresting. (This is why astazangasta's criticism in another comment is so amusing - this is meaningful because the sample is so 'biased'. This also holds for all the other studies done on Framingham.) But because only one highly homogeneous sample is studied, the chance genetic similarities of friends aren't due to just similar ethnic backgrounds, but are being caused by the genetic influence on things like SES and personality and intelligence and religiosity and hobbies and things like that. These similarities would probably also hold when considering a more unusual and cosmopolitan group, but it would be difficult to spot the increased similarity on a genome-wide basis because of all the racial differences in genomes (most of which would be nonfunctional and irrelevant to anything).

a_bonobo• 10 years ago

They only apply population stratification from Figure 2 including onwards:

The results so far do not control for population stratification because we wanted to characterize overall similarity. However, it is important to remember that some of the similarity in genotypes can be explained by simple assortment into relationships with people who have the same ancestral background. The Framingham Heart Study is composed of mostly whites (e.g., of Italian descent), so it is possible that a simple preference for ethnically similar others could explain the results in Fig. 1. However, in the following results, we applied strict controls for population stratification to ensure that any correlation we observed was not due to such a process.

I don't understand why they did not do this for Figure 1 (or why they do not show that figure) - I assume that the difference disappeared, but who knows.

I just realized that this paper is from 2014, first published on arxiv in 2013. I can't find any study replicating it in the 34 citations since so I would be careful (but this type of dataset for replication is hard to come by) https://scholar.google.com.au/scholar?cites=7301900347564089...

microcolonel• 10 years ago

I live in Toronto, and frankly I doubt my Panjabi, Sri Lankan, Chinese, and Korean friends are as genetically similar as my fourth cousins.

eggie• 10 years ago

It could be that you share certain alleles at a much higher rate than expected by chance.

That said, if the study is working off the Framingham data, I do wonder how general a conclusion it is...

rogerthis• 10 years ago

Haven't read the text yet, but the title remembered me of the Hungarian psychiatrist Leopold Szondi.

astazangasta4• 10 years ago

This is a dataset designed to study genetics in a cohort of mostly white people from Framingham, MA. How can anyone claim to be able to draw conclusions about the genetics of friendship from this cohort? This is such a biased dataset I can't believe this got published.

For the record, I am Indian, my best friend growing up was Italian, my current best friends are a Russian Jew, a Scotch Irish mix from Woburn, and a Canadian of some uncertain European temperament. I have never had a best friend even close to being my fourth cousin.

Manishearth4• 10 years ago

They explicitly call this out and apply population stratification later in the paper.

astazangasta4• 10 years ago

Yeah, which means the extensibility of these claims is exactly as good as those corrections, which can't possibly hope to model the missing information (the actual genetic distribution of friends).

qq66• 10 years ago

The genes that you get from your ethnicity are a small fraction of your overall genome. There are tons of other genes that affect things like your height, your interest in music, etc, that could strongly affect your friend selection.

astazangasta4• 10 years ago

And I got those in some other way than my ancestry...?

popmystack• 10 years ago

I'm not sure you understand the whole genetics thing.

kbd2• 10 years ago

Pretty sure my Thai best friend is more dissimilar to me than a fourth cousin.

awl1302• 10 years ago

unless you are thai yourself? are we supposed to presume that everyone is not-thai on this site?

kbd2• 10 years ago

Fair enough. I'm not even Asian.

sndean• 10 years ago

There's a pretty good CGP Grey video that explains family trees and what fourth cousins are:

https://www.youtube.com/watch?v=PM79Epw_cp8

LoSboccacc• 10 years ago

* fourth cousins genetically as diverse as random strangers would be a better wording

the current title has implications not yet rigorously tested and people is already looking into selection bias to explain correlation where probably there's none.

tomtomtom777• 10 years ago

fourth cousins genetically as diverse as random strangers would be a better wording

That is not the same at all. From the abstract:

More than any other species, humans form social ties to individuals who are neither kin nor mates, and these ties tend to be with similar people.

They specifically tested friends to include the general preference of friends to be similar.

compactmani• 10 years ago

The main finding from my perspective is that we can score and rank friendship likeliness and see signal in a hold out dataset.

I can almost hear Facebook clamoring to offer free genotyping.

Snoooze2• 10 years ago

The dating world is already experimenting. (e.g. http://www.genepartner.com/)

cafard• 10 years ago

How many Americans know any fourth cousins? In the village from which two great grandparents emigrated, I have met some third cousins, some at least twice removed, and here in the US I have met second cousins once removed. For what it's worth, it would take a very sharp eye, and maybe an unduly imaginative one, to spot family resemblances between the third cousins and me.

tomrod• 10 years ago

I know plenty, due to both an active family reunion group and due to genealogical research. It can be pretty fun learning about your ancestors.

avodonosov4• 10 years ago

Can it be interpreted that friends are as similar as random people?

Snoooze2• 10 years ago

No. The article mentions a number of reasons why friends can be (and are) genetically closer than random people.

jlebrech• 10 years ago

so what would be the best way to increase genetic diversity?

monkpit• 10 years ago

Why are you under the impression that it needs to be increased?

return02• 10 years ago

Why would we need that. Assuming we 've gotten over our survival issues a looong time ago, what would be the benefit other than weirdly looking humans?

neolefty• 10 years ago

Study abroad!

thesz2• 10 years ago

Third cousins are most fertile.

So, marrying friend is a good idea from fertility point of view.

amelius• 10 years ago

"Fertility" does not necessarily mean "fit offspring", but your point stands.

thesz2• 10 years ago

Define "fit offspring".

I consider "offspring that has many offsprings" as one that fits definition "fit offspring". If you read the article you would know that third cousins' couples have more grandsons. Thus, third cousins direct offsprings are... more fit than others!

TheLarch4• 10 years ago

The lead author, Nicholas Christakis, is the man who was reviled by some students at Yale during the Halloween costume debacle: http://www.theatlantic.com/politics/archive/2015/11/the-new-....

sliverstorm• 10 years ago

I hope you're not suggesting that has any bearing on the quality of the paper?

eric_h2• 10 years ago

I don't think there's any implication of that. It's just a nice piece of trivia.

bmelton• 10 years ago

From my reading, the article in question gives me a complimentary view of the author, and a less complimentary view of those who reviled him.

TheLarch4• 10 years ago

Not sure if you are speculating that I am a supporter or detractor of his. My only impression of the man up until now was that I thought he and his wife conducted themselves with grace and forebearance during the Halloween fiasco. I find this research fascinating.

eric_h2• 10 years ago

Thanks for reposting that article. I swear I'm starting to feel like an old man "back in my day, vigorous intellectual disagreement and debate was the point of living on campus during college"

kafkaesq• 10 years ago

Back in my day, vigorous intellectual disagreement and debate was the point of living on campus during college

Well they've sure fixed that, haven't they.

TheLarch4• 10 years ago

[TheLarch proffers curmudgeon fist bump]

peter3033• 10 years ago

Most people with European ancestry are 10th or 11th cousins to each other based upon how every US president is related to every other. This includes the current President who is half Eoropean.

Retric3• 10 years ago

2 ^ 11 = 2048 great ... grandparents. However, the top row only counts as one set. Thus, at most 1,024 couples. 10 generations down with 3 kids a generation is 59,049. Assuming zero deaths and zero intermingling that's 60,466,176.

So, your going to need very large family sizes for that to work out once you account for deaths without children or intermingling. France averaged 4.5 kids a couple in 1800 so I don't think this math works.

peter3033• 10 years ago

Its not math, but exhaustive geneologies by experts who've tracked every US President as far back as Saxon England. These emperical results constrain theoretical calculations.

dragonwriter• 10 years ago

Most people with European ancestry are 10th or 11th cousins to each other based upon how every US president is related to every other.

"The people who have been elected President of the United States" are not a representative sample of "people currently alive with European ancestry" (or even a nonrepresentative subset of that set) such that generalizations about relations between members of the latter class could be drawn from the relations between members of the former class.

And even if they were, strong conclusions can't really be drawn from an N=44 sample.

aj7• 10 years ago

Very weak correlations. Spun as a hypothesis-positive result.

peter3033• 10 years ago

Assuming you have 44,000 genes (a copy from each parent) and random recombination each generation, then some of your ancestors will have stoped contributing genes by 16 generations (400 years). These assumptions arent perfect, but givebyou a sense of what happens.

(People have far fewer than 2^N ancestors due to distant cousin inbreeding) (Genetic recombination is not random, but chunky.)

Retric3• 10 years ago

This is not necessarily accurate as a fair amount of what we think of as Junk DNA is actually used. And a large chunk of DNA is devoted to basic cellular function which fails if it's modified.

bencollier49• 10 years ago

Doesn't 23andme "identify" fourth cousins?

awl1302• 10 years ago

in other words, we pick friends from the same ethnic group

Friends are as genetically similar as fourth cousins

Related Stories