In a recent Panda's Thumb comment thread, Pam asked (among other things) about our human species genetic Adam and Eve:
I have been reading for the last few years now, that there is a consensus among the majority, that humans have been genetically traced to a two human ancestory: A genetic "Adam and Eve".
This is a relatively common misconception, and a very understandable one. There have been published studies that have looked at the most recent common mitochondrial DNA ancestor of all humans, and other studies that have looked at the most recent common Y-chromosome ancestor of all humans. Since mitochondrial DNA is only inherited from the mother, the most recent mitochondrial DNA ancestor is frequently referred to as "mitochondrial Eve." Similarly, since the Y-chromosome is passed on exclusively from father to son, the most recent Y-chromosome ancestor gets called "Y-chromosome Adam" a lot. The use of those two terms is not entirely inappropriate, but it can be very misleading - particularly for those who haven't taken a bunch of college-level biology.
Let's start with the biggest misconception, and move on from there. When it comes to Adam and Eve, there's good news and bad news. The good news is that I can unequivocally state that they never got divorced. The bad news is that they never married. That's understandable, of course, since Eve died more than 50,000 years before Adam was born.
Right about now, I'm guessing that we're hitting the point when confusion sets in. After all, if every man on earth is descended from Y-chromosome Adam, and if we're all descended from our fathers, and if Y-chromosome Adam was married, why wasn't his wife "Eve?" We all clearly must be descended from her, too, right? (And if confusion hadn't set in already, it almost certainly has now that I wrote that.) Let's see what (if anything) I can do to clear things up.
It's probably a good idea at this point to step back for a minute or two and review some really basic genetics.
When we talk about a gene, we're usually referring to a specific stretch of DNA. Each stretch of DNA has a location on a chromosome. Humans (with some rare and typically pathological exceptions) have 46 chromosomes - two copies of each of the 22 autosomal chromosomes, and two sex-determining chromosomes (women have two copies of the X-chromosome, and men have one X-chromosome and one Y-chromosome). When we reproduce, we each give our children 23 chromosomes - one of the two copies of our autosomal chromosomes, and one of our two sex-determining chromosomes.
We also have mitochondria in our cells. These are little organelles that are a key part of the process that lets us use oxygen to make energy. Mitochondria have their own DNA, and reproduce on their own within our cells. With very, very rare exceptions, we get all of our mitochondria from the egg, which means that we only inherit mitochondrial DNA from our mothers. (So, guys, you can feel free, the next time your wife complains about our unequal contribution to childbirth, to point out that we contribute less than half of the DNA, so it's only fair that we also do less than half of the work. If you do this, have your next of kin tell me how it went.)
That's the genetic basis, and the easy part. Now we get to how genes are passed on, and here's where it gets complex.
What it comes down to is this: we get half of our chromosomes from each parent, but that doesn't mean each of our grandparents contributed 25%, or that each of our great-grandparents chipped in 12.5%. If you go back far enough, you will find that the majority of your biological ancestors have left you with absolutely nothing in the way of DNA. That's where this gets really fun - at least for the mathematical population biology geeks.
It's possible to think of each one of your genes - each single copy - as if it was an independent organism. A tiny little thing that has a parent and that can have offspring. That's a fiction, of course, but a useful one. If we use that as our model, we can trace the lineage of a single gene instead of a single person. It's a really convenient way of doing things (partly because the math gets a little easier).
If we treat genes as individual organisms, we can create a "gene genealogy" - a gene family tree - for individual genes. Here's an example I just made. Since I'm more or less completely unskilled when it comes to developing complex simulations with a computer, I raided one of my kids' board games. I started with five individuals, and rolled a die for each. On a 1 or a 2, the individual produced no offspring. On a three or a four, it had a single offspring. On a five or a six, it produced two kids. Here's the result:
As you can see, after four generations, all of the individuals shared a single ancestor for that gene. But not all genes are going to have the same gene genealogies. In fact, hardly any will. Let's look at another example, using exactly the same rules:
In this scenario, not all individuals shared a single ancestor for this gene after four generations, and it's clear that this gene is going to have a very different genealogy from the first. In fact, this time it took a lot longer before all of the offspring in a generation shared a common ancestor for this gene:
It took much longer to arrive at a point when all of the members of a generation shared a common ancestor, but when that finally happened, the most recent common ancestor wasn't part of the initial generation - it was a member of generation 5.
It is possible for two very different gene genealogies to exist within a single population. Remember, we get one copy of each gene from each of our parents, and they get one copy from each of theirs. That means that if your maternal grandmother contributed the DNA that determines your blood type (for example), then your maternal grandfather didn't play any role in your blood type.
This diagram is a bit more complex than the last one, but it's the best I could do (at least at present). It represents a hypothetical gene genealogy for your blood type. You are the yellow triangle. Biological descent is shown with lines, and descent for the one gene - let's call it "eyebrow thickness" - is shown with the colored arrows:
Now, let's do the same thing, but look at a different trait - let's call this one "nose length":
If you look at the two gene genealogies superimposed on each other - I'm going to save your eyes, and let you do that in your imagination - you will see that:
- Your maternal grandmother is your biological ancestor and your "eyebrow thickness" ancestor, but not your "nose length" ancestor.
- Your maternal grandfather is your biological ancestor and your "nose length" ancestor, but not your "eyebrow thickness" ancestor.
- Your paternal grandmother is your biological ancestor, your "nose length" ancestor, and your "eyebrow thickess" ancestor.
- Your paternal grandfather is your biological ancestor, but not your "nose length" ancestor or your "eyebrow length" ancestor.
- Your maternal grandfather does have biological descendants, but he does not have "eyebrow thickness" descendants.
As you can see, different genes can have different genealogies even within a single biological genealogy. That's why it's not only possible but also unsurprising that the mitochondrial gene genealogy that traced back to "mitochondrial eve" and the Y-chromosome genealogy that traced back to "Y-chromosome Adam" don't lead back to individuals who were married to each other, or even lived within 50,000 years of each other. It's just part of the glorious complexity that you get to see when you look at the way genes get passed through populations.
The real reason that you hear about "mitochondrial Eve" and "Y-chromosome Adam" instead of "nose length Edith" and "eyebrow thickness Archie" is because of the way that those particular markers - the mitochondria and the Y-chromosome - are passed from generation to generation. Unlike most of our genes, these have very, very simple genealogies. Mitochondria are passed on from mother to daughter, and the mitochondrial lines end whenever you have a woman who has only sons (or no kids at all, of course). Y-chromosomes are passed on from father to son, and the Y-chromosome lines end whenever you have a man who has only daughters (or, again, no kids at all). For this, and other reasons, they are much easier to work with for these sorts of studies. And, of course, you have the added bonus of knowing that the last common Y-chromosome ancestor was a man, so you know you can get away with calling him "Adam" - after all, he's been dead for so long that we'd never know that he was really named "Steve."
- Log in to post comments
The link from PT to this post is dead.
Yeah, I forgot to put the link in. I hate it when I do that. It's fixed now. Thanks.
A very nice, succinct, easy-to-understand explanation. Thanks for posting it.
"who has only sons (or no kids at all, of course)" = has no daughters
"who has only daughters (or, again, no kids at all)" = has no sons
Wonderful post -- I had never thought about the genealogies of individual genes before. Good stuff!
Great post, Mike. I have some questions:
Why is mito-Eve older than Y-Adam? If today's parental patterns are any clue, women get children earlier than men so one should think their generation time might be, say, 25 years and men's, 30. If this pattern (not the numbers) holds out for a long time, one would think that it would take less time for mitochondria to spread in the population than for Y-chromosomes?
Or is it because there are less genes on the Y chromosome than in the mitochondrial genome? Is this last bit even relevant?
Torben:
This is because a dominant male sometimes corrals nearly the women in a group and fathers most of the next generation, cutting off a number of Y chromosome lines and replacing them with his own.
You can work out the implications if you draw a few diagrams like Mike's.
Excellent explanation of something I haven't understood before. As a math teacher interested in the whole question of chains of different sorts - causal steps, probability trees, etc., I really appreciated your diagrams and explanation.
Torben: Because there's greater disparity in reproductive success in men than in women. Even an extremely successful woman won't have more than ten or so children survive to produce grandchildren for her. However, a male can conceivably produce hundreds of offspring and thousands of grandchildren. And over many generations, this makes it relatively more likely that some man's descendants will be a large portion of the population than any particular woman's.
Good post. I would like to point out that a creationist organization called Reasons to Believe exploits the Y-chromosome "Adam" and Mitochondrial "Eve" concept to claim that we descended from Noah's family, and further back, Adam & Eve, respectively. They are wrong for the same reasons that it would be wrong to say you had only two grandparents because you only got one gene from each of two for eye color. Because they leave out everything except for Y-chromosome and Mitochondrial DNA, they mislead their followers into thinking that no one else existed at the time. I'll certainly cite this post when I write about them on my own blog. Thanks!
For a wonderful example of male reproductive success, see this paper. As a fellow I know said upon reading this, "Whatever else you may say about him, Genghis Khan was a genuinely fit individual."
Insert standard disclaimer here about not actually proving that it was Genghis Khan, merely an amusing idea, etc., etc.
Another incredibly well-done explanation.
Excellent post. Keep up the great work.
Boo Yah! (Quite possibly the first time an ESPN expression has been used to descibe the excellence of a science post.)
So, if I'm reading this right, in addition to Mitochondrial Eve and Y-Chromosome Adam, there should also exist a Nose-Length Edith and, say, Freckles Pete, it's just that it would be much more difficult (if not actually impossible) to track them down?
If that's the case, would it be possible to construct some sort of meaningful minimum set of common ancestors, a small group of ancient people who could be taken as the source of all the traits we have today?
Or is it too complex and too spread out for that to be at all meaningful?
Pete, Scott: Thanks. I guess Genghis Khan proves you right.
One point you might wish to address: I've heard some trying to "prove" the literal truth of the Old Testament through DNA studies argue that the "Y-chromosome Adam" is actually Noah. Since he and his three sons were the only human males on the ark, only his Y chromosome would have been present after the flood. Therefore, a biblical literalist should expect the genetic male ancestor-of-all to live significantly after the genetic "Eve."
This doesn't explain the 50,000 years between "Eve" and "Noah," but someone committed to a <10,000 YO erth isn't going to buy that part of the argument anyway.
Another way of conceptualizing the problem is to think of family names, which are traditionally passed by male descendants (in the US). For example, consider the Amish: They were founded by a number of families yet in some places you'll find a predominance of names like "Zook" or "Stoltzfus". It's not that there weren't other Amish families living and intermarrying in the community all along, it's just that the Zooks and Stoltzfuses happen to have founded lineages that produced the greatest number of boys who: A) Lived long enough to have children and B) Never failed to have at least one reproducing male in each generation to carry on the family name.
Thanks Mike, an excellent explanation, the 'gene geneology' diagrams were very helpful.
One question though, how can we know the last common mitochondrial ancestor without analysing the mitochondrial DNA of every female or of a known ancestor of every female?
The same applies to men for the last common Y chromosome ancestor.
The Seed overlords really need to a "Print This" button capability to Sb. This post is really print worthy!
Yes, thanks. Several of us tried to answer Pam reasonably, but it's impossible to explain very well in a short post, and I doubt I could do so well even at the length you used.
I'll probably be linking here if the issue comes up again.
Glen D
http://www.geocities.com/interelectromagnetic
lmf3b,
The problem with the biblical literalists is that they ignore the rest of the data. If there were 8 people on board a big boat (Noah and family) that resulted in all of us, then every allele would converge on those 8 people, who have 5 effective genomes. Additionally, everyone's mitochondrial DNA would converge to just 4 or less different mitochondrial ancestors at the exact same time.
Reading further... if i have this right, Noah and his thee sons only had male offspring... where did everyone's mitochondria come from then? Geez, this is what happens when you use an old myth to try to make predictions about reality.
The take-home lesson is that, exactly as Ross suggests, every gene and its alleles has a common ancestor (or CONCESTOR as Dawkins coined the term) akin to Y-chromosome Adam and Mitochondrial Eve. To say that everyone descended from "Freckles Pete" is absurd, as it is to say the same for the Y/Mito ancestors.
Neil-- I believe the initial effort to track back to the mitochondrial Eve was done by Allan Wilson and Rebecca Cann at Berkeley. Although they obviously didn't test everybody, they tried to get blood samples from a wide variety of ethnic groups, including Australians. There has been a lot of work on this topic since. Cann and Wilson's original conclusions-- which didn't get universal acceptance when they came out 20 years ago-- have apparently been basically confirmed.
Another point to make is that there are other "named" common ancestors other than Y-Adam and MT-Eve. There is like a HLA-John (or something, I need to look it up).
[younger brother's mind explodes]
Just FYI, when making flat-colour diagram images like that on a computer, you really don't want to use the JPEG file format, which is designed for compressing photos and results in ugly artifacts on other types of images. The PNG format will produce both better-looking images and smaller file sizes (although only if saved from non-JPG masters!); see http://www.r1ch.net/img-formats/
Interesting. I always assumed it was mostly down to population size. Roughly twice as many people with mitochondria as with Y chromosomes, after all.
Reed's comment on HLA would be difficult to figure out. We do have multiple HLA genes and these form haplotypes, but the genes themselves have complex histories and selection has probably played a big role in keeping some alleles in the population. There is recombination between homologous genes that would combine two lineages, and there is evidence of gene conversion where different copies of the various gene family members within a haplotype may have recombined combining lineages created by the recombination between genes in the haplotype etc.. It would be a mess to figure out.
The Y chromosome and mitochondrial DNA are easier in this regard because very little recombination between lineages takes place.
Ron Okimoto
It would seem to me the picture should be a lot fuzzier because the DNA strands do not sit about like rocks but are actively involved in breaking and recombining. Supposedly that occurs at a constant rate, but like dice, there is no memory, so it might be fast one week and slow another, and only average at a given rate.
If one looks at diversity, one can make an informed judgment about (but only that)about how long such bouncing around occurred and thus the dates given. As you point out the transfer is still in discreet packets but there are many more than two possibilities.
The two possibilities you mention are the most likely (maybe?)but in any gene the possibility of crossover exists (as well as errors) so you could inherit a-n of one grand parent and n-z of the other even on the same chromosome.
So while you might give a general estimate of the time frame involved (give or take a dozen generations)you can neither identify the location or specific details about that person as it would be like reading a book where all the pages had been switched about.
The other math would entail building a simple genealogy chart and presuming a name for each label. As the number of places on that chart doubles with each generation the number of labels exceeds the world population in four hundred years or so, so pretty quickly a single name shows up on a lot of labels, but it would virtually never show up on ALL of them.
In short if you had a complete turnover every 400 years or so where each person was descended from every possible ancestor then the very idea of a single "Eve" even for Mitachondrial DNA seems far fetched as the exact age of any particular DNA change cannot be gathered, and dozens if not thousands of people could have indistinguishable DNA from that distance in time, in that a change might occur in a dozen generations or already exist.
In paired Genes it is much trickier(and the male gene is paired at least in part.)So again an "Adam" as some sort of breeding genius is a comforting thought but mathematically impossible.
"If you look at the two gene genealogies superimposed on each other - I'm going to save your eyes, and let you do that in your imagination"
But apparently you did superimpose them -- the two diagrams are identical.
As for why Y-Chromosomal Adam is more recent than Mitichondrial Eve, it follows from a greater statistical variation in the number of offspring men have.
the very idea of a single "Eve" even for Mitachondrial DNA seems far fetched
You're confused. Consider the number of women alive today. Consider the number of mothers of women alive today -- it cannot be more, and will be less whenever two women share a mother. So the number of female ancestors steadily decreases as one travels up the tree, eventually reaching one as a matter of logical necessity.
Interesting. I always assumed it was mostly down to population size. Roughly twice as many people with mitochondria as with Y chromosomes, after all.
That's irrelevant, as all male offspring are a dead end in regard to inheritance of as mitochondria. The mitochondrial tree is a tree of mothers just as the Y chromosome tree is a tree of fathers.
Greater statistical variation in the number of offspring in men vs. women means broader levels in the tree of survivors, and greater breadth means less depth. It's basic graph theory.
To say that everyone descended from "Freckles Pete" is absurd, as it is to say the same for the Y/Mito ancestors.
It's not absurd, it's true by definition: everyone is descended from each of their ancestors and, assuming common descent, there is some common ancestor of everyone with mitochondria, of everyone with a Y chromosome, etc.
The Seed overlords really need to a "Print This" button capability to Sb. This post is really print worthy!
Click on "view source" in your browser, select everything from 'In a recent' through 'he was really named "Steve."', paste that into a file with an extension of .html, open that file in the browser, and print it.
One question though, how can we know the last common mitochondrial ancestor without analysing the mitochondrial DNA of every female or of a known ancestor of every female?
How can we know anyone who has been dead for over 100,000 years? We don't know her, we just know a little bit about her -- approximately when she lived, and that she was our common mitochondrial ancestor.
I've heard some trying to "prove" the literal truth of the Old Testament through DNA studies argue that the "Y-chromosome Adam" is actually Noah. Since he and his three sons were the only human males on the ark, only his Y chromosome would have been present after the flood.
That Noah might have been Y-chromosome Adam -- or rather, that some facts about him aren't inconsistent with that -- doesn't prove that he was. Unfortunately, most human beings, not just Biblical literalists, have a very poor grasp of basic logic.
If today's parental patterns are any clue, women get children earlier than men so one should think their generation time might be, say, 25 years and men's, 30.
Even if there were any evidence in support of the claim that fathers are generally older than mothers, it would have no bearing whatsover on the shape of the descendent tree -- how could it?
If this pattern (not the numbers) holds out for a long time, one would think that it would take less time for mitochondria to spread in the population than for Y-chromosomes?
Huh? Everyone has mitochondria, and all men have Y-chromosomes; talk of differential rates of spread is nonsensical. The issue here isn't about anything spreading through the population, but about the relative shapes of two trees, one with all men as leaves and one with all women as leaves.
Or is it because there are less genes on the Y chromosome than in the mitochondrial genome? Is this last bit even relevant?
Of course not; how could it be? The specific characteristics of Y chromosomes and mitochondria are completely and utterly irrelevant here. All that is relevant about them is that we get Y chromosomes from our fathers and mitochondria from our mothers.
Thanks hoary puccoon, from your info I've been able to find out that in the original research they sampled the mtDNA of 147 people and how they used this to estimate the age and geographical location of the last common mitochondrial ancestor.
How can we know anyone who has been dead for over 100,000 years? We don't know her, we just know a little bit about her -- approximately when she lived, and that she was our common mitochondrial ancestor.
tm, I think you are taking 'know the last common mitochondrial ancestor' just a bit too literally there. I was meaning 'know when/where she lived', I wasn't really expecting her personal interests and phone number. Thanks for the other comments though, explaining why Y-Chromosomal Adam is more recent than Mitichondrial Eve.
I can't believe no-one's put in the I'd-have-thought-obligatory Stephen Fry quote, here. From the very first episode of his masterful series "Q.I. (Quite Interesting)":
"... but perhaps, you know, we should believe in Adam and Eve... Geneticists have established that every woman in the world shares a single female ancestor who lived a hundred and fifty thousand years ago. Scientists do actually call her "Eve". And every man shares a single male ancestor dubbed "Adam". It's also been established, however, that Adam was born eighty thousand years after Eve... so the world before him was one of heavy- to industrial-strength lesbianism, one assumes."
TM
I do take your point of mitochondrial DNA being more conservative for the lack of a dance partner, but 400 years, much less 20,000 is a very long time in terms of random happenstance.
In any case the entire thing is an arithmetic exercise only, as, like checking footprints after the passing of a herd of cattle, only tiny bits remain, and those bits would be widely held in common in any case.
My main point however is that gene pools are just that, and while individuals vary, they are each decedent of the entire pool, including such leakages as existed between pools.
A man might look forward a thousand years and discover that everyone alive was descended from him at least once, but almost everyone else could do the same thing.
One point that I don't think has been mentioned is that it's a little weird to be talking about Eve as some specific, interesting person...since our Eve can change from time to time.
If she had two daughters (she must have had two or more - if she only had one, then she wouldn't be Mitochondrial Eve, her daughter would), it's almost certain that one of them has more extant descendants than the other.
If all except one living woman descended from Daughter A, and one woman descended from Daughter B, then if that woman died childless, or if she only had sons (who would die eventually), then there would no longer be any living descendants of Daughter B - thus making Daughter A Mitochondrial Eve.
In short, every time someone dies, there's a chance that the actual hominid we call Mitochondrial Eve would cease to be Mitochondrial Eve, and that her daughter would inherit that title.
I think it was Dennett who made this point in...some book or other.
Since Genghis Khan has already been mentioned, I will mention another highly "fit" historical male- Edward III of England, who has been called the "Father of the English Middle Class" (through his many, many illegitimate children) and according to some estimates has millions of descendants in America alone.
My personal favorite concept in the human genealogical tree is the "Identical Ancestors Point", the point at which every human then alive was either a) an ancestor of all humans currently alive or b) an ancestor of no humans currently alive.
One of the consequences of this is that humans have a mitochondrial Eve (& the Adam), but you'll also find "Eves" further back. There might be an mitochondrial Eve for the apes (or apes plus humans, depending on your terms), a mitochondrial Eve for monkeys plus apes, and there might also be a primate mitochondrial Eve.
I bring this up because of the study done on cats. Unfortunately, I haven't been able to read the paper, but it appears that they found an "Adam" and "Eve" for the cats, yet they also track strains of cats by more recent "mitochondrial Eves". That's what the secondary reports indicate (I think the primary source is supposed to be Science, perhaps on the web and in the future, paper).
Just an expansion on what "mitochondrial Eve" and "y-chromosome Adam" can mean. They can exist around various levels of taxonomy, or show up where no divergence occurred, like with our most recent "mitochondrial Eve".
Glen D
http://geocities.com/interelectromagnetic
As I pointed out that could easily be in as little as 500 years if there was any travel at all, or in genetically isolated groups only a couple hundred. Easter Island, and Pitcairn Island are famous examples as well the group in Maracaibo who had one visitor with Huntington's disease and now all are her decendents, half with the gene and half without it.(also decended of all the other folk then as well.)
In 10,000 years even such odd occurrences as a boat blown off course, or the odd traveler would connect otherwise isolated groups, as the newcomer would bring the entire population of his home country of a few hundred years before that.
Likewise a European shipwreck off Newfoundland might well have decendents in Patagonia eventually even though there was no more travel of more than a hundred miles at a generation, and no recording of the link.
Sorry, but i don`t trust....
Outstanding post!
This is a bit old but some may be interested in it:
http://gregladen.com/wordpress/wp-content/pdf/LadenNakedTruth.pdf
Surely this is all much easier to explain using a few basic queries? For example, are we all exactly the same species and should all new species of life start with a small population of a single pair of unique genes/DNA? Do all species go extinct? Are near-extinction events regular enough to restart the process? I can understand that 'Adam' and 'Eve' may have been tens of thousands of years apart but were they exactly the same species of homo sapiens sapiens? Men are just simply not as successful at passing on their genes as women because in theory a great great grandfather of 60 years old (possible) can have kids with many other mens great-grand-daughters and skip 3 or 4 generations?