Genetic genealogy can reveal a fascinating picture of your racial origins - but how reliable is the data for the family historian?
By Megan Smolenyak
Last updated 2011-02-17
Genetic genealogy can reveal a fascinating picture of your racial origins - but how reliable is the data for the family historian?
You've seen it on television and read about in the newspapers, and it makes you think twice. The head of Harvard University's African-American Studies department takes some DNA tests and is startled to learn that both his paternal and maternal lines are European, and that he has as much European ancestry as African. An adoptee who always thought her family originated from South America takes a test and is pleased to learn that she's of Native American ancestry. Just what is 'genetic genealogy' and is there anything you could learn from such testing?
Simply put, genetic genealogy is DNA testing that's done specifically for the purpose of learning about one's heritage. There are a number of myths about the topic, and one of the most prevalent is the notion that it's the same kind of testing that's done for medical or criminal purposes. It's not.
The DNA tests used by family historians are more restricted, and in a sense, more innocent, than those you see on shows such as 'CSI'. If police wanted to use these tests to find suspects, they'd be forced to round up hundreds or even thousands of distant cousins. And although some wish they would, these tests do not shed any light on your predisposition for certain medical conditions.
Surname studies are by far the most popular application, but many are unaware that there are a handful of tests available, and depending on what you're curious about, one of the other options may be of greatest interest to you. For those who aren't keen on needles, the good news is that all commercial testing companies use mouth swabs, so the whole process is blissfully painless.
If you're already a family historian enthusiast, you may well ponder whether it's worth it. After all, the most popular tests run about £75-100. If any of the following apply, you probably should consider adding DNA to your genealogical arsenal:
As you would expect with any technology, new types of testing continue to emerge, but we'll briefly explore the most popular types available today, starting with the granddaddy of them all, Y-DNA.
Y-DNA testing uses the Y chromosome, which is only sported by males of the species. This has contributed to a myth that women can't play. Just as many take their first steps in genealogy by googling their surname, most venture into genetic genealogy by testing their surname - and I was no different. So when I wanted to test the family I was born into, I picked up the phone and asked my father to take the test. If he had refused, I could have turned to one of my brothers or a paternal uncle or cousin.
The reason Y-DNA is so irresistible to family historians is because it's passed intact from father to son down through the generations.
The reason Y-DNA is so irresistible to family historians is because it's passed intact from father to son down through the generations. This is wildly convenient since surnames are passed the same way in most cultures. Barring a non-paternity event (yes, they do happen, so make sure you can live with the results if they surface a surprise in your family tree!), the surname and Y-DNA travel through time in tandem - meaning that every man walking the planet today carries the same Y-DNA genetic signature as his father, grandfather, great-grandfather, etc. He's a living representative of those who came before, so no - and this is yet another popular myth - it's not necessary to dig up the dearly departed.
Most of genetic genealogy boils down to a matchmaking game, and Y-DNA is no exception. When you receive your results, it will likely look like gibberish - just a bunch of numbers. Since it's still cost-prohibitive to test our entire genetic makeup, these tests rely on selected markers, which can be thought of as landmarks in the landscape of our DNA. Genealogists have piggybacked off the efforts of population geneticists who have identified markers that are highly variable - ones that are useful for distinguishing between peoples and even individuals. Your results will be presented as a number for each one of these markers, and it's these numbers (representing how often certain genetic patterns repeat themselves) that are used for finding matches.
This kind of testing can't tell you that your most recent common ancestor is your mutual great-great-great-grandfather, but you know for sure that your lines converge at some point.
When you use a conventional genealogy database and enter 'John Smith,' the system looks for matches for each letter: J-o-h-n . . . This is exactly what happens with genetic genealogy, only numbers are used as the basis of comparison. Fortunately, all commercial testing companies have databases that automatically generate a list of matches for you. Provided you signed a release, you can then communicate with your genetic mates. Assuming you've joined one of the thousands of surname studies that already exist (try entering the following terms into a search engine - 'genealogy, DNA, surname' to see if there's a project for your name), you'll want to look for matches among your fellow participants. If you want to extend the search still further, you can enter your results in free, public access databases such as www.ybase.org, www.smgf.org and www.ysearch.org and look for matches.
So what does a match mean? It means you share a common ancestor with that person. This kind of testing can't tell you that your most recent common ancestor is your mutual great-great-great-grandfather, but you know for sure that your lines converge at some point. In most surname studies, clusters of genetic mates emerge, and these folks become your research pals. If you're especially lucky, you'll find yourself in a cluster with others who have already done a lot more homework than you. I recently persuaded a man named Jim Shields to join the Shields project of which I'm a member, and he matched another participant perfectly. The other fellow had researched the family for 27 years and knew exactly where Jim fit in, so Jim suddenly had 450 new relatives! And while I can't say this is an every day experience quite yet, the databases are finally achieving critical mass - that is, there are now enough of us getting tested that this kind of outcome will become increasingly common.
If you are contemplating a Y-DNA test, I would suggest that you select at least a mid-resolution test - a test with more than 20 markers. While tests with fewer markers are less expensive, they too often lead to false positives. You may match someone perfectly at 10 or 12 markers and both upgrade to 43, only to learn that you now only match on 33 of those 43 markers. In other words, your initial match was misleading.
After tiptoeing into the world of genetic genealogy with a Y-DNA test, many become curious to see what else they can learn, so they move on to mitochondrial DNA. Many think of mtDNA as the maternal version of Y-DNA, and while there are certain parallels, it's important to understand the differences - starting with the mode of transmission. MtDNA is passed from mothers to both their sons and daughters, but the sons don't pass it on. This is fortunate because it means all of us - male and female - can take this test ourselves. As with Y-DNA, our test represents those who came before us, only in our maternal line - our mother, mother's mother, mother's mother's mother, etc.
...recent research has revealed that there's greater variety in mtDNA genetic signatures ... and this suggests that it will be of greater genealogical value than we had imagined.
But at this point, mtDNA is not quite as genealogically useful as Y-DNA. This is because it's largely been considered more of a deep ancestry test. If you're familiar with Dr. Bryan Sykes' book, 'The Seven Daughters of Eve', you're aware that the basic premise is that 95 percent of those of European origin can trace their maternal roots to one of seven women who lived between 10,000 and 45,000 years ago. When you take an mtDNA test, you're learning which one of these 'daughters of Eve' you descend from (and incidentally, there are about 36 or 37 on a global basis) - and most companies will provide a color map showing roughly how and when your branch of the world's maternal family tree (referred to as your 'haplogroup') migrated out of Africa.
That's wonderful to know from an intellectual curiosity standpoint, but it usually doesn't tell you much about your recent roots. My maternal Irish forebears, for instance, have passed the H haplogroup down to me. H happens to be the most common in Europe (apparently her maternal descendants were the most successful in reproducing, so roughly 30-40 percent of Europeans are also H), so I have literally millions of maternal cousins.
Fortunately, however, recent research has revealed that there's greater variety in mtDNA genetic signatures ('haplotypes') than previously thought, and this suggests that it will be of greater genealogical value than we had imagined. Also, a few companies have recently introduced full sequence mtDNA tests, so in the not-too-distant future, the matchmaking game for mtDNA may well be as effective as it is with Y-DNA.
MtDNA also has a role in dealing with specific genealogical conundrums. For instance, if your great-grandfather had 15 children by three wives, and you can't quite assign all the children to the correct mothers, you might be able to test direct line maternal descendants of a few of the daughters involved and back-door into their respective mothers. It takes a little strategizing, but it's possible! MtDNA is the tool that scientists most frequently rely on for history mystery situations, such as the recent identification of the Titanic baby as Sidney Leslie Goodwin of Fulham, west London.
Should you opt to take an mtDNA test, I suggest seeking one that covers both HVR1 and HVR2 - 'HVR' standing for hyper-variable region. Testing only HVR1 will produce low resolution results and probably more matches than most would be interested in investigating, whereas testing both regions will result in more precise results and a narrower field. Of course, if you've got deep pockets, you can go ahead and test your entire mtDNA sequence, but you'll be waiting for a while to play the matchmaking game since this is so new that very few have taken such tests.
'Ethnic tests' isn't a formal category of genetic genealogy, but it's a convenient way to refer to a variety of tests that mostly reveal something about one's geographic or ethnic origins, such as African, Viking, Cohanim (Jewish priesthood) and Native American tests.
In general, it's wise to be slightly more wary of these tests than others because it's not realistic to expect our DNA to neatly map to present-day countries or peoples. It's true that there are some fairly strong correlations between certain genetic signatures and ethnic groups (Cohanim comes to mind), but these associations can be hard to interpret and there will always be exceptions. DNA has been around far longer than national boundaries, and mankind has been on the move and intermingling for millennia, so it's a bit of wishful thinking to look for tests to definitively classify your paternal line as Italian or Estonian or Welsh.
Some geographic or tribal affiliation can be provided based on where one's genetic mates reside in Africa today.
Having said this, there's a place for such tests, provided you understand the limitations. As with the tests already covered, ethnic tests do not explore your entire family tree and genetic makeup. Most come in a Y-DNA or mtDNA version (or both) and test only one branch of your family tree.
And your results may surprise you. In the case of African ancestry tests, for example, approximately 25-30 percent of those tested will learn that their paternal line is European in origin. This is the legacy of plantation society coming down to us in our genes, and was seen in BBC's 'Motherland: A Genetic Journey'.
In the other 70-75 percent of cases, some geographic or tribal affiliation can be provided based on where one's genetic mates reside in Africa today. But Africa is the cradle of mankind, so we've had longer there to move around than on any other continent - so the notion that one's maternal or paternal genetic signature will be found in only one location in Africa today is somewhat optimistic. A recent American television series, African American Lives, involved the testing of nine celebrities, and as the results were revealed, almost all discovered that they have genetic cousins in several locations in Africa. So while it's a definite improvement over what had been available previously to learn about one's African origins - that is, virtually nothing - the results should be taken more as an indication than an absolute.
Another entertaining option is the BioGeographical or admixture test. This one breaks your geographic heritage into four broad chunks: European, African, Asian and Native American. I'm a blue-eyed blonde, so I was somewhat surprised by my results: 86 percent European, eight percent Native American and six percent African. At least, that's what's been calculated as being the most likely scenario. Since this test involves Y-DNA, mtDNA and autosomal DNA (the kind that's randomly shuffled at each birth event), there's probability involved. And when you look at my results, you see that the range of possibility for both my Native American and African heritage includes zero. For instance, my most likely African results are given as six percent, but the range shown in the company's report goes all the way from zero to fifteen percent. My European tops out at a maximum of about 96 percent, so it seems I have some admixture, but I'd be overstating things if I were to announce to the world that I'm 86/8/6.
Due to the use of autosomal DNA, this test is also somewhat shallow - meaning that after so many generations, the input of any given ancestor essentially washes out. So someone taking the test to prove African ancestry from the 1700s, for instance, will probably find that the genetic echoes have diminished to the point where they simply don't register.
This is an amusing test, but as with the ethnic tests, one should take the results as more of an indication than an absolute. It's also worth mentioning that for those who test as at least 50 percent European, there is now an advanced EuroDNA test that can break that portion of your heritage down more finely. And that's an example of one of the best aspects of genetic genealogy - it keeps moving forward. I, for one, can't wait to see what they might be able to reveal a few years down the road!
There are many different companies offering genetic services - a selection can be found below. It's best to ask around before committing to any company - magazines, forums and online communities are a good place to start.
Trace Your Roots with DNA: Using Genetic Tests to Explore Your Family Tree by Megan Smolenyak Smolenyak and Ann Turner (the best-selling how-to book on genetic genealogy)
The Seven Daughters of Eve by Bryan Sykes (an engaging introduction to the world of mitochondrial DNA)
The Journey of Man: A Genetic Odyssey by Spencer Wells (a fascinating exploration of Y-DNA and man’s migration across the globe)
Blood of the Isles: Understanding Our Tribal History by Bryan Sykes (a highly readable look at genetic patterns across the British Isles)
This page is best viewed in an up-to-date web browser with style sheets (CSS) enabled. While you will be able to view the content of this page in your current browser, you will not be able to get the full visual experience. Please consider upgrading your browser software or enabling style sheets (CSS) if you are able to do so.