The Genetics of Genealogy

Tweet about this on Twitter0Share on Facebook0Share on Reddit0Email this to someone

In many ways, we feel disconnected. … I think this is simply the latest application of technology to allow people to connect with the past that, in many cases, they feel that they’ve lost. … I think it is part of a larger movement to connect with our perceived past, who our ancestors might have been, because we feel that we’ve lost that.

Spencer Wells on Open Source

Click to Listen to the Show (24 MB MP3)

Where’s your branch? [Maia C / Flickr]

We’ve long been a nation of amateur genealogists. We trace roots. We draw family trees. We hang our grandfather’s clan’s tartan next to our grandmother’s Navajo rug. Largely a nation of immigrants, voluntary or forced, it’s how we make sense of how we got here and who we are. (Which is why, I guess, genealogical websites rank second only to porn on the internet.)

This is the parallel, personalized history of the U.S., a history not of the country’s Founding Fathers but of our own founding fathers — and mothers. And whether the Old Country was Portugal or Poland, Ghana or Guatemala — or perhaps all of them — there’s a good chance that someone in your family is filling in a particular branch of our wonderfully mongrel American tree.

Until recently, this would have meant poring through Ellis Island logs, town clerk’s records, maybe slave ship manifests, and, increasingly the Mormons’ mammoth International Genealogical Index, the largest genealogical database in the world. And, of course, taking part in an oral tradition: listening to and passing on familial creation truths… or myths.

Now comes genetic analysis, which promises to add some meat to the bones of your cherished Mayflower stories — or smash them to bits. For anywhere from a little over $100 to $1000, you can now find out all kinds of things about your history. Do you want a migration map of your likely American Indian forbears? Might you have Jewish ancestry? (Henry Louis Gates does, he recently found out.) You could go for the Cohanim/Levite test and move up a level in your temple’s ritual hierarchy. Or figure out how many generations back you have to go to find out the most recent ancestor you share with your next-door neighbor, who seems like she couldn’t be more different than you.

Is this something you’d want to do? Is this something you’ve already done? What did you learn? What did you do with what you learned? When the molecular evidence contradicts the family lore, do you still print the lore?

It’s clear what we gain when we add the certainty of DNA analysis to our stories and sense of who we are. But what do we lose?

Spencer Wells

Explorer-in-residence, National Geographic

Project leader, The Genographic Project

Author, The Journey of Man

John Hawks

Blogger, Johnhawks.net, an anthropology weblog

Assistant Professor of Anthropology, University

of Wisconsin–Madison

Andy Carvin

Blogger, Andy Carvin’s Waste of Bandwith and DNA Cousins

Director, Digital Divide Network

Y-chromosome Haplogroup J; Mitochondrial DNA Haplogroup pre-HV


  • http://www.scienceblogs.com/gnxp razib

    I have posted about this in detail here, Finding out who your “ancestors” were via DNA. Here is my conclusion:

    These scientists and companies are preying on human needs and intuitions which are being inappropriately triggered by these findings (part of it is semantic, part of it is reasonable extrapolation based on what I would call “folk genetics”). The reality is that our ancestry is extremely reticulated, it does not extend indefinitely outward back into genetic space in an every expanding mushroom cloud of twigs, it begins to wrap back onto itself and twists into a ball of spaghetti which eventually becomes attenuated via founder events (the “Out of Africa” one being the most prominent). A Y chromosomal or mtDNA lineage will tell you only so much, and more so on the populational level (the soup of alleles can perhaps give you clues to the migrations between populations, at least that’s the theory). Even the more nuanced autosomal tests, which sample many loci on the genome instead of just a lineage through one nonrecombining marker, are problematic in their framing. South Asians routinely get back results that suggest they are admixed, but that’s only because the tests are geared toward Europeans, East Asians and Africans. The preponderance of the evidence is that South Asians are not admixtures at all, but their own populational cluster, with various relationships to other geographic races. The further and further we move back in generations the less and less our conventional intuitions matter, who we are as individuals becomes diffused into thousands of humans who carry bits and pieces of information which will eventually make it into us. But, selection also matters, alleles, genetic variants, may sweep across populations without an concomitant replacement of other regions of the genome. In the deep past pedigrees are irrelevant, bodies are nothing but way stations where collections of genes get together for a generation.

    In short, it’s complicated. But the scientists who oversell this work and the companies selling you ancestry-in-a-test-tube would not benefit much from admitting this.

  • http://www.scienceblogs.com/gnxp razib

    Or figure out how many generations back you have to go to find out the most recent ancestor you share with your next-door neighbor, who seems like she couldn’t be more different than you.

    most of the tests do not find your most recent common ancestory, they find your most recent common uniparental ancestor. that is, mtDNA tests find the mother’s-mother’s-mother’s-mother’s-mother’s-mother you share with your neigbhor. they would miss your mother’s-mother’s-father’s-mother because the uniparental mtDNA lineage is passed only through mothers. the same applies to Y chromosomal tests, they might catch your father’s-father’s-father’s-father’s-father’s-father’s-father’s-father’s-father, but they would miss your father’s-mother’s-father. there are promises of getting around this by using regular autosomal DNA (stuff that’s not passed by sex).

    these tests do tell you something, but far less than they hint at.

  • http://www.scienceblogs.com/gnxp razib

    to be short: these tests won’t tell you something you don’t know unless you are in special circumstances, eg., an adoptee which no knowledge of their forebears, or a black american who wants to get an idea of the region of africa some of his forebears came from. there will be white americans who will register as “minority asian,” but this is almost certainly statistical noise. south asians are always labelled a mix of races, but that’s because the tests are geared toward testing the “major” races (even though south asians are 1/6 of the world’s population we aren’t a “major” race).

  • http://www.scienceblogs.com/gnxp razib

    also, re: guests, my friends henry harpending is a genetic anthropologist at u of utah, or john hawks, u of wisconsin biological anthropologist, would offer skeptical voices in this area. if you want pro-genetic testing guests, spencer wells, bryan sykes, mark shriver and stephen oppenheimer.

  • http://www.scienceblogs.com/gnxp razib

    also, i think there is a difference between testing one snip of you genome, and sequencing the whole genome. the latter could give you some real information, because it gives you a full image rather than a tiny slice. and, we might have $1,000 dollar personal genomes soon.

  • eugene_X

    There is a very eye-opening section in Richard Dawkins’ newest book, “The Ancestor’s Tale,” which deals with this exact topic. Specifically, with a television show aired by the BBC a few years ago in which a woman of British/Jamaican heritage was sheperded, via DNA tests, back to her “ancestral tribe” in West Africa.

    Dawkins rips this apart quite handily, for reasons similar to those mentioned above, and really criticizes the premise of the show, for promising a degree of certainty which in Dawkins’ view, is simply dishonest. I can’t explain in any more detail, as I have loaned out my copy of his book, and I don’t want to paraphrase it inaccurately, but it is well worth having a look at it.

  • http://www.oroboro.com/rafael rafael

    Get ready for a whole new resurgence of racism.

    At the start of the 20th century, “scientific” racism and eugenics was mainstream science. Since that time the idea of races as a scientific classification of people has been demolished. It has been determined that the everyday concept of the races, who is white, who is black etc. is too inexact to mean anything or predict anything. The catch phrase is that race is a “social construct”.

    And since in practice the current concept of race can’t predict anything, this in turn gives a rational support to the moral argument that society should not discriminate by race.

    That support is about to give way. As these tests become mainstream people will be able to classify themselves in genetic clusters with scientific precision. The boundaries of the “races” may change, but the new boundaries will be much more exact and much more clearly defined.

    Let me give you a few examples of the new racism:

    1. There was that study recently by Gregory Cochran about how Ashkenazi Jews may be more intelligent than the general population. His argument is a based on certain genes which affect the brain and their relationship to certain genetic illnesses among that population – like Tay-Sachs. Another would

    2. Herrenstein and Murray’s “The Bell Curve”

    I don’t want to comment on the validity of the study or the book ( I have read neither ), but rather to point out that studies like this are now possible and will become more common. This will have vast social implications.

    We are already dealing with issues of genetically based discrimination for health insurance. You can make a moral argument that everyone should be allowed equal access to health insurance, but you can no longer make an argument that it is not rational for the insurance company to want to discriminate.

    What will we do if in fact we are able to conclusively tie certain performance charcteristics to specific genes? Even just the objectively measurable ones – like reaction time, physical strength, or resistance to cold – let alone more subjective ones like intelligence or creativity. Will there be pressure to start regulating which people can get which jobs? Will we start adjusting SAT test scores to normalize out genetically based ‘deficiencies’? Will someone with a specific cluster of genes which does not currently map onto a recognized “race” be able to claim discrimination based on evidence that people with such a cluster are under represented in some given context?

    BTW I’m don’t mean this to sound like I’m in support of the revival of the destructive concept of racism, but just that access to new information about genes and what they do will bring back difficult debates about race which we will not be able to easily dismiss.

  • http://www.scienceblogs.com/gnxp razib

    Dawkins rips this apart quite handily, for reasons similar to those mentioned above, and really criticizes the premise of the show, for promising a degree of certainty which in Dawkins’ view, is simply dishonest. I can’t explain in any more detail, as I have loaned out my copy of his book, and I don’t want to paraphrase it inaccurately, but it is well worth having a look at it.

    ok, consider the cohen modal haplotype (CMH). this is a Y lineage very common in the kohanim, the jewish priestly caste whose status is passed from father to son. the presence of the CMH has been used to diagnose the likely jewish origins of the lemba of zimbabwe (at least in part), the authencity of jewish ancestral claims by the “crypto-jews” of new mexico who are now identified as catholic hispanics. but, the rub is this, not all cohen carry this marker, and not all who carry the marker are cohens. this speaks to the issue dawkins brings up in regards to the certainty of ancestrally informative markers, they are informative, but, they are not determinative. cohen or yoruba status is in large part socially mediated, so even though genetics can give us broad outlines it can never impart certitude. if an african american woman is told her ancestors were yoruba because she exhibits a yoruba modal mtDNA, that neglects the reality that neighboring groups also might carry these haplotype, and some yoruba do not carry it. granted, among african americans the information they have on hand isn’t much, so it is better than nothing, but it needs to be weighed and interpreted properly.

  • http://www.scienceblogs.com/gnxp razib

    but you can no longer make an argument that it is not rational for the insurance company to want to discriminate.

    who has ever made the argument that it is not rational for insurance companies to discriminate? that is why some people pay higher premiums! that is why some people are “insurable.” that is why gov. passes laws which prevent insurance companies from following the perfect implications of their acturial tables. if someone tells you they were admitted into harvard, that tells you a lot about their cognitive capacity, but how does that influence how your treat them?

    What will we do if in fact we are able to conclusively tie certain performance charcteristics to specific genes?

    here is the problem with your argument, a gene A is not always (usually) either necessary or sufficient for trait X. many traits we are alluding to, like disease susceptibilities and intelligence, not only have environmental/developmental components, but, they are polygenic. these traits and their genetic clusters are easy to speak of on the group populational level, that is, you can ascertain about how much of the populational variation is due to genotypic variation, but, you can’t always speak much to the individual level. if you are going to use insurance actuarial tables then this information is valuable, but, government does not operate on probabilistic procedures, it is deterministic, and these genes are usually not going to be deterministic.

    to make it short: you might be able to calculate a heightened probability of crime as an expectation if you have cluster of genes a,b,c,d,etc., but due to variance there won’t be definitive predictive power.

  • http://www.oroboro.com/rafael rafael

    Razib:

    It is always rational for the insurance company to discriminate based on actuarial tables. That is to discriminate based on observed probabilities of socially constructed groups – like the current definition of race. Say for example that the insurance company finds that people of African ancestry are more prone to heart disease and wants to adjust their tables. ( One can always object to this morally of course, but leave that aside for now ) An individual can object to having their rates raised by claiming that the race of an individual cannot be determined exactly, and that they themselves are not in fact part of the affected group. African ancestry is a perfect example of this, because especially in America, people of African ancestry are very diverse genetically.

    But for a highly specific genetic test there is no such appeal. If you have certain alleles your personal increased suceptibility to certain illnesses can be predicted very exactly. And this is just the tip of the iceberg, because over time we will have a greater and greater ability to make these predictions.

    As to your second point, many disease suceptibilities are in fact caused by mutation in a single gene such as Huntington’s disease and cystic fibrosis. But even in the case of polygenic traits, like susceptibiliy to Alzheimers or diabetes, one can make highly accurate predictions about likelihood of outcome based on the presence of a particular allele. That is, there is nothing difficult about taking a population of people, looking at disease outcomes, and measuring the likelihood that a particular allele predicts susceptibility to a particular illness. In fact this is a common avenue of research. And as the cost of genome mapping comes down, and computing power goes up, our ability to make these kinds of predictions is increasing rapidly.

    So far this kind of knowledge has been confined mostly to medical research. I point out that in the last 10 years this new information has started to be used for non-medical arguments, most commonly in discussions of intelligence. Up until recently arguments about intelligence vs. race have involved a lot of handwaving about who is in a particular group. And since the groups are socially constructed, arguments about differences in intelligence have been confounded by social factors like unequal access to education, class, cultural bias etc. Recent advances in genetic testing however make it possible to eliminate these factors. For example I can look at the presence of a gene predicting intelligence within a particular social group. For example I can look at just among asians, from a certain city, and a certain socioeconomic background, and compare intelligence vs. certain alleles. I might even do a population study comparing just brothers, and measuring the intelligence of each vs. the presence of certain genes. This kind of study would make a powerful argument about intelligence vs. genes which is impervious to attack based on social factors like access to education or class.

    Even if this information is confined to the hands of scientists who are properly trained in statistical analysis it will have profound implications for society.

    Now with the advent of these new “geneology” and “race” tests, we are going to find this information getting into the hands of people with their own racial ideologies and who don’t even know how to interpret this kind of information – but who will be armed with what looks like ‘incontrovertible scientific proof’, which will look to the average person just as convincing as the sorts of arguments coming out of the scientific community.

    Consider for example this kind of information in the hands of congressmen. *shudder*

  • http://www.scienceblogs.com/gnxp razib

    But for a highly specific genetic test there is no such appeal. If you have certain alleles your personal increased suceptibility to certain illnesses can be predicted very exactly.

    allele A in genetic background Z is not the same as allele A in genetic background Y (at least not necessarily). on acturial grounds the arguements works, but on technical-scientific grounds there still is fiddle room because modern medicine has a large grey space of epistatic interactions which are just being explored. even though we can speak with rough statisical knowledge, the biological roots of this knowledge are still filled with large gaps and assumptions. insurance companies can rationally make an argument for discriminating based on probabilities, but, on moral-fundamental grounds the biological grounding is not as rock-hard as the probabilities on a populational levle might imply.

    But even in the case of polygenic traits, like susceptibiliy to Alzheimers or diabetes, one can make highly accurate predictions about likelihood of outcome based on the presence of a particular allele. That is, there is nothing difficult about taking a population of people, looking at disease outcomes, and measuring the likelihood that a particular allele predicts susceptibility to a particular illness. In fact this is a common avenue of research. And as the cost of genome mapping comes down, and computing power goes up, our ability to make these kinds of predictions is increasing rapidly.

    1) i agree that these studies are going to be more accurate

    2) but, these gene association studies are not as simple and clear cut as you seem to imply (from my reading of the human genetics literature at least).

    3) the likelihood functions one generates can be used on non-genetic parameters too, so in a probabilistic fashion there is nothing novel here.

    4) so my point is that the smoking gun (which is in sight, but no here yet) will be a biological understanding on a deep level, which the probabilities don’t necessarily give you.

    as for this:

    This kind of study would make a powerful argument about intelligence vs. genes which is impervious to attack based on social factors like access to education or class.

    this is a complicated topic. but if you want me to be frank, yes, i think genes for IQ that are differentially represented in different groups will be found, and some, like ASPM, seem to have been found. my main issue with your commentary is that you are eliding the distinction between probability and determinism. the two are very different. also, a minor point is that empirically predictive models on a probabilistic level don’t necessarily reflect biological reality in any exact sense.

  • http://www.oroboro.com/rafael rafael

    razib,

    Your point is well taken – I am not really making a strong distinction between increased likelihood vs absolute certainty. But I don’t think this distinction is necessary for my point. I’m saying that the ability to make performance predictions about people based on genetic tests with scientific certainty will create challenging questions for society. By certainty I don’t mean that the prediction must be determinstic, but rather that one can have a high degree of confidence in the prediction whether the prediction is determinsitic or as is more likely just an increased probability.

    For example, right now when a school looks at a population of children they have no way of knowing who will do well at math and who won’t. This is comforting in that it encourages everyone to treat the children equally, and also it gives each child the feeling that they have the same chance as any other child at being good at math. But we may soon live in a world where we can tell with complete certainty that certain children are more likely to be better at math. E.g. you can have a person in front of you and know for sure that they come from a small population of people who are say 20% better at math. And for each child you may be able to classify them into a small group whose range of math performance is somewhat predictable. The important thing is not that the prediction is determinstic, but that a prediction of some kind can be attached to a particular person with certainty. This will work against the desire to treat people equally.

  • http://www.scienceblogs.com/gnxp razib

    For example, right now when a school looks at a population of children they have no way of knowing who will do well at math and who won’t.

    actually, psychometrics might not be perfect, but this is the whole underpinning (now mostly implicit and tacit because of the ‘forbidden’ nature of IQ testing) of standardized tests: you can predict from test performance general cognitive aptitudes. the fact is that these tests and socioeconomic parameters are extremely good predictors on a probabilistic level. the probability that a kid from brookline with parents who are science professors and scores a 140 on raven’s matrices culture-free tests at age 6 is going to complete calculus is far higher than a kid from southie with high school drop out parents who scores an 80 at raven’s matrices culture-free at age 6. our rough knowledge is there, the genetics is simply another layer of precision.

    The important thing is not that the prediction is determinstic, but that a prediction of some kind can be attached to a particular person with certainty.

    i am a bit confused about what you mean by “predict with certainty.” probabilities are conditional, the more information you have the more you can predict with confidence, but predictions have errors or variances attached to them. basically, you seem to be implying that genetic testing is going to reduce the error involved in the likelihood functions, and this is probably so, but it will not eliminate the error. ultimately you question goes back to issues of the perception of free will, etc.

  • http://www.oroboro.com/rafael rafael

    Razib:

    By predict with certainty I mean that I can be certain about the correctness of my prediction. E.g. If I have two bags A and B. Bag A has 5 red and 4 black marbles, the other has 4 red and 5 black. I can be completely certain that there is a higher chance of pulling a red marble out of bag A than bag B. I may not be able to predict what marble will get pulled from each bag, but I have an iron clad argument that each bag gives different odds.

    I agree that one can be very successful at predicting academic success by looking at socio-economic factors. Currently as a society we believe that if we swapped those children at birth, that the social factors would still explain the performance difference. And while it is currently plausible to say that there may be a genetic component in performance disparities, there has been up until recently no practical way to test this – which is good because it encourages us as a society to work toward erasing disparities, and to invest the same effort into every child for example.

    Soon we will be able to say with some certainty exactly what our genes predict about our future. Perhaps the predictions they make will be so vague as to not be practically useful – but early evidence says otherwise.

    The difference between predictions based on social factors vs. genetic factors, is that we generally believe that its a good thing to change those social factors to make society better and fairer. Attempting to change genetic factors in the same way would be the return of ‘eugenics’.

  • http://www.scienceblogs.com/gnxp razib

    If I have two bags A and B. Bag A has 5 red and 4 black marbles, the other has 4 red and 5 black. I can be completely certain that there is a higher chance of pulling a red marble out of bag A than bag B.

    where your analogy is problematic for me is that you can look inside the bags, A & B, and know a priori what the probabilities of drawing red & black marbles are for respective bags. in regards to biological examples we are discussing here the probabilities are derived a posterior from inferences about the bags. of course, sequence level data is ironc clab, but we don’t care about the sequence except insofar as how it relates to phenotype, and i think that the mapping is not as clear cut as you might imply.

    please understand, i think a lot of the new genomic science, and i think it can tell us a lot. but, as a classical liberal i think we are making too much of it if we think it will overturn legality under the law, because

    1) legality under the law is an a priori assumption

    2) the law is premised on deterministic dynamics, not probabilistic ones

    Currently as a society we believe that if we swapped those children at birth, that the social factors would still explain the performance difference.

    i disagree. a segment of the elite has promoted this, and so it is normative in the media, but there have been dissenters in the academy and on the popular level a debased genetic determinism is very popular.

    And while it is currently plausible to say that there may be a genetic component in performance disparities, there has been up until recently no practical way to test this

    there are ways to test this. thomas bouchard’s twins separated at birth studies and comparisons of fraternal and identical twins (same environment, different genetic endowments) are rough and ready ways to test this. there are issues with heritability calculations, but it isn’t like they haven’t been out there.

    Perhaps the predictions they make will be so vague as to not be practically useful – but early evidence says otherwise.

    the predictions aren’t going to be vague, but, they will be better by a difference of degree rather than kind. behavioral genetics has been around before behavorial genomics, the latter is a sharper and more refined tool, but the same arguments used against behavioral genetics can be trotted out against behavioral genomics.

    i think the issue we are talking about is psychological, not scientific. genomics will be a new tool in assessing conditional probabilities with greater predictivity, but the thing is that the public perceives it to be a new sort of magic. once it becomes banal i suspect that the specter of dystopia will fade.

    and yes, we will start reshaping out genome soon.

  • http://www.scienceblogs.com/gnxp razib

    also, let me be clear here: i think genomics is a revolutionary thing, the amount of data we have is mind boggling. but, some of the things that rafael alludes to have been around for a long time. we could differentiate populations a long time ago theoretically, we just needed more loci. lewontin’s 85% variation within population vs. 15% variation between population was true, but only for one locus. analysis of multiple loci could elucidate ‘correlation structure’ and allow population clusters to be separable pretty easily. an analogy is trying to classify races by hair color alone, there is a lot of intrapopulational variance, but if you take hair color, straightness, thickness, etc. you can differentiate people much easier. eg, someone with black kinky hair is often african, someone with thick straight black hair is likely to be east asian, etc. etc. the more different data points you have the better picture you can construct. the principles have been around for a while, but now we have more data.

    in regards to how we view society, life and the meaning of it all, i think cognitive science and neurobiology are going to be much more world shattering than the new genomics, because the new genomics is a predominantly statistical science. many of the traits we are interested in have multiple causative parameters, there won’t be one smoking gun on an individual level, even if we get a good grip on it on a populational level.

  • skfranz

    I used to work in a genetic breast cancer clinic and these questions were very much a part of our protocol. For an individual who has already had breast cancer the risk of genetic testing being a factor in loosing insurability is minimal — but for someone else in the same family it could be a lifesaving test and could result in insurability. Eventually we’re all going to be found to be carrying a dozen or more genes with increased risk for a dozen or more serious diseases — one of which we may die from. We could all be “cherry picked” out of the insurance pool with enough information available since we’re all going to die of something.

    On a more personal basis — one line of my family is, by tradition, Native American but there are a lot of questions about the time period and some of the family members involved. There is ancedotal evidence that some of them were shunned by other white family members for marrying outside southern white society. However, at least to me, it seems equally likely that those marriages might have been African American considering the time and place. The kind of testing that’s being done now could be a pointer to us as to where to look for more information, though it might (or might not) be surprising to a few of my cousins.

  • skfranz

    Sorry — correction. “Could result in a loss of insurability.”

  • http://www.scienceblogs.com/gnxp razib

    well, i suspect that genetic testing will probably lead to universal coverage because too high a proportion of the population (middle class) is going to feel insecure and will be terrified out of the current system. kind of off topic :)

  • avecfrites

    This is fascinating stuff, and I’m sure it will help us better understand how people came to populate various land masses and all that PBS stuff. But somehow other images overpower that — I see a variety of packaged distractions, wedge issues, home DNA testing kits, and the new phrenology. Who needs skin color and regional accents anymore; now we can classify you scientifically, digitally, with color bar charts and no appeals.

    I’d like to see people take the DNA test and get a uniform response: you’re an American through and through, you have the right to be here, and here’s your official user guide and certificate of authenticity, called the Constitution.

  • tbrucia

    Here’s some interesting generational math… If one has two parents, four grandparents, eight great-grandparents, etc, then the formula for the number of direct ancestors of the same level (A) is A=2^n. If you set n equal to 33, and apply this formula, you get 8.589 billion great-great-etc whatevers. Wow, that’s a lot of ancestors!!!! Next step, figure each generation at 25 years (a rough guesstimate, but in the neighborhood). Now, going back 33 generations takes you back 825 years (33×25), or the year 1181 AD. Yep, the population of the world back then — and now — isn’t 8.589 billion. So basically we are all subject to an awful lot of our direct ancestors jumping in the sack with other of our direct ancestors. Makes one realize there’s more truth than fiction in the idea of the human family… and a somewhat incestuous one at that!

  • http://www.andycarvin.com andycarvin

    Hi everyone,

    As it turns out, US News did a cover story about me on this very subject five years ago; I was one of the very first customers of FamilyTreeDNA.com to get interesting results. They matched me with a guy in Philly, and US News had us do a family reunion; I also turned out to carry the so-called Cohen gene. Since then, they’ve managed to connect me to over 40 men with DNA so similar, we probably share a common ancestor within the last 300 years. Weirdest of all, one of those 40 men turned out to be my friend and colleague, Ethan Zuckerman. I’ve spent the last five years learning everything I can about population genetics in references to my particular Y chromosome and mtDNA patterns. Fascinating stuff….

  • http://www.andycarvin.com andycarvin

    Update: turns out I was undercounting the number of men who’ve been matched with sharing my particular y chromsome. The list is now up to 70+ men, from as far away as south america. I’ve also set up a DNA Cousins blog for members of this extended genetic family to post updates about their family tree research, as well as news stories on DNA and population genetics. I may also do the same for my mitochondrial DNA group, but it’s much smaller. My mtDNA, known as pre-hv1, is generally quite rare. It probably originated in Yemen and is common in Arab populations, the horn of Africa, Iran, even the Kalash tribe of Pakistan.

  • serious lee

    I recently checked out my jeans and found out I’m related to Levi Strauss.

  • http://www.scienceblogs.com/gnxp razib

    mad props to john. he might not be as pretty as spencer, but he knows his human genetics. at the current rate john you’ll become such a big public intellectual that you’ll be guaranteed tenure denial!

    a question for spencer, last year you were interviewed in discover where you addressed the reality of human variation. and recently you offered nick wade some quotes for his commentary on the idea of recent human cognitive evolution, which does have a human variation angle. are you going to look at functional differences after the neutral lineage gig is exhauste?

  • nother

    When I see people engage in these searches in more than a moderate way, I come away with the feeling that this can be another symptom of our deep yearning for immortality. We tend to want to stretch this short life out either way, future or past. It’s very important to focus on these precious fleeting years we have.

    I have a feeling some of the people searching deep into their genealogy are coming away with more knowledge about their distant ancestors than the knowledge they have about their living relatives. I bet when they find one of these names from 200 years ago they say to themselves, wouldn’t it be great if I knew all about this person, how they lived, what they thought? Guess what, you can know more about the family around you, call up uncle Joe, your second cousin Sue, have that in-depth conversation. 200 years from now one of your future relatives will be doing a computer search on your second cousin Sue and will be asking themselves, wouldn’t it be great if I had a chance to know more about this person.

  • http://www.scienceblogs.com/gnxp razib

    I am not Richard Dawkins!

  • http://www.scienceblogs.com/gnxp razib

    Your are welcome Chris.

  • terence

    Dr Wells

    Going back 60,000 years is quite a jump. Perhaps we would be more focused if we asked questions like:

    Christopher Columbus just had an audience with the Queen . Where were your genes?

    The Normans are invading and Harold must defend us! Where were your genes?

    This general has crossed the Rubicon, and is heading to Rome. Where were your genes?

    Get them big stones over here by Tuesday, or we will not get this Stonehenge thing up and running until next week!

    Where were your genes?

    I am an avid amature genealogist. I am familiar with family lines going back about 300 years.

    I have a particular interest in cousins that marry cousins.

    I have put together a handfull of charts demonstrating that a child has only four great grandparents, instead of the eight that we would normally expect.

    Also those lineages that demonstrate that one person alive in 1995 has five distinct paths back to a woman alive in

    1650.

    When the DNA of these persons is examined, how will it differ from the DNA of a person who has six generations of ancestors

    that were NOT cousins?

    Does the multiple connections to a third great grandfather increase the chance of a certain marker appearing?

    I can see that the number of ancestors, as you said on the PRI program, does not double each generation, but increase by a factor of 1.4 or 1.8 over 200 generations.

    Other than the two markers discussed (M & F) do you see in the future the ability to trace OTHER markers?

  • Potter

    We’ve decided to partake in the National Geographic Genographic Project. Too enticing. Will not get involved in more than a very moderate way Nother, promise.

    I may be related to Andy Carvin.

  • Nikos

    It’s probably too late to catch anyone’s attention now, but I just read an analysis of the ‘Cohanim geneology’ in Jonathan Marks’s What it Means to be 98% Chimpanzee @ http://www.amazon.com/gp/product/B00006AG20/002-7919812-6315255?v=glance&n=551440

  • serious lee

    Hello

  • Pingback: conditional license class amature radio