Demystifying DNA Tests

By Megan Smolenyak Smolenyak Premium

You probably first heard about it a few years ago, but now it’s everywhere. Your society is sponsoring a lecture, fellow researchers were chatting about it at the Family History Center, or maybe you got an e-mail asking you to participate in a surname study. But you’re unsure about jumping on the DNA bandwagon. How does genetic genealogy work? How much does it cost? What about privacy? And why bother, anyway—will it really tell you something that traditional genealogy can’t?

Genetic genealogy complements traditional research—it’s not a substitute for squinting at microfilmed records and scouring family histories. You can’t simply take a DNA test, plug the results into an online database and discover your whole family tree back to the Dark Ages. The odds of finding a meaningful match in a DNA database aren’t particularly strong, since not many people have participated yet. And of course, DNA testing won’t reveal your ancestors’ names and birth dates. But it can help by confirming (or disproving) family legends and research discoveries. You can use it to find out if you’re distant cousin to other Kowalskis, to determine if the Colonial-era Massachusetts and Virginia Austin clans are connected, or to learn the truth behind that yarn about Granddad being adopted. Ultimately, genetic genealogy can save you time and money in researching the old-fashioned way.

That might sound counterintuitive if the $100 to $300 DNA-test price tag gives you sticker shock. But I became a believer when I conducted my own surname study. All Smolenyaks trace their roots to one of four families from a town in present-day Slovakia. But the paper trail peters out in the 1700s, so we couldn’t find a relationship. Imagine my surprise when DNA testing revealed that none of the families shared a common ancestor. My disappointment wore off, though, once I realized I’d been spared years of frustration trying to prove a false belief. DNA testing can steer your research by telling you which paths to avoid and hinting at where you should be looking. Fortunately, you don’t have to be a scientist to tap into this new genealogical resource. We’ll explain how genealogy-DNA testing works, how safe it is, and what it can (and can’t) tell you about your family.

It’s in Your Genes

If you’ve browsed DNA testing companies’ Web sites, you’ve undoubtedly noticed a wide array of tests. Which one’s right for you?

Your genealogical goals will tell you. Let’s take a look at the possibilities.

Y-DNA (or Y-chromosome): No doubt about it — Y-DNA testing is king. It wasn’t the first test, but it’s the most popular due to the Y-DNA-surname connection: Only men have Y chromosomes, which are passed essentially unchanged from father to son — quite convenient, since surnames also are passed on this way. As shown in the “Peripheral Visions” box below, all the men in the top line of a pedigree chart share the same Y-DNA — as do all the male-line descendants of an ancestor perched on top of a descendancy chart. Because Y-DNA and surnames travel together through time (with the occasional exception of a non paternity event such as illegitimacy or adoption), Y-DNA tests are handy for surname projects that determine whether families with the same last name are, in fact, related. And it’s not necessary to dig up ancestors’ graves, since a man alive today sports the same Y-DNA as his paternal-line great-great-(insert as many greats as you’d like)-grandfather.

Two, 20 or 200 men named Johnson can get tested, and those who share an ancestor will have identical (or nearly identical) Y-DNA results. Women can participate, too, but not directly: Since females don’t have Y chromosomes, they have to persuade a father, brother, uncle or cousin to take the test.

Typically, a surname project starts when an enthusiastic genealogist invites others with the same surname to test and compare results. (To see if your last name is one of the 1,000 plus that are being studied, peruse <>.) Frequently, eight or nine people join a freshly launched project, and no matches appear. Just as frustration is about to set in, the next participant matches someone. Eventually, clusters of matches emerge.

Each matching group shares a set of Y-DNA marker values. I like to think of markers as landmarks in the landscape of our DNA. Scientists test certain markers that are useful for distinguishing between people and populations. Your test results are presented in alleles — a series of numbers that indicate how often certain genetic patterns (called short tandem repeats, STRs, or just repeats) recur at each marker tested. You can order Y-chromosome DNA tests ranging from 10 to 43 markers — the more markers, the more precisely the test can indicate how many generations ago you and your match share a common ancestor. A 43/43 match, for instance, is more convincing than a 12/12 match. Two men conceivably could match each other on all markers from a 12-marker test, but only 35 markers on a 43-marker test. As you’d expect, the more markers you test, the more you pay. Many people opt for the middle ground and choose tests with 21 to 37 markers.

Although each testing company has its own preferred set of markers, they overlap quite a bit, so it’s usually easy to compare test results from different companies. Collectively, your marker values form your haplotype. To see if we belong to the same haplotype, you and I would simply compare our numbers.

For instance, two men with similar surnames might wonder if they share an ancestor. Testing can’t reveal who a common ancestor is, but it can confirm or disprove a connection and estimate roughly when the most recent common ancestor (MRCA) lived. Let’s say that both men take 25-marker tests and wind up with a near match. They have identical results, except that one has 31 repeats on marker 449, while the other has 32 repeats. The two men have slightly different haplotypes — it looks like they’re related, but there’s been a mutation. Mutations serve as a generational clock: The more mutations, the longer ago the MRCA lived. So the results indicate that the men have a common ancestor, but their MRCA probably lived longer ago than the MRCA of two men who match perfectly.

Mitochondrial DNA (mtDNA): Many think of mtDNA as the maternal version of Y-DNA, and there are parallels, but it’s important to grasp a few differences. Mothers pass mtDNA to both their sons and daughters, but sons don’t pass it on to their children. Just as Y-DNA traces the top line of a pedigree chart, mtDNA follows the bottom line, traveling from mother to daughter. Once sons pass away, their mtDNA lines die out.

For genealogy, mtDNA isn’t as useful as Y-DNA because it’s a “deep ancestry” test, involving time frames that are more anthropological than genealogical. You may have heard of Bryan Sykes’ The Seven Daughters of Eve, a book based on the premise that 95 percent of people of European origin can trace their roots to seven women who lived 10,000 to 45,000 years ago. Scientists have tested populations worldwide, and so far, they’ve uncovered about 36 “daughters of Eve” — each is akin to a maternal branch of the world’s family tree. The mtDNA test reveals which of these daughters you descend from, and expresses that result by assigning you to a haplogroup. All members of a hap-logroup share the same genetic mutation as their “daughter of Eve,” but their haplotypes may differ from each other depending on additional mutations that have occurred over time.

My haplogroup is H, the most common European type. That gives me some insight into how and when my maternal line migrated to and within Europe, but it also means I’m distant cousins with more than 40 percent of Europeans, and we all share a maternal ancestor who lived approximately 20,000 years ago. From a genealogical perspective, that’s not especially helpful.

But mtDNA can assist in your roots pursuit in three noteworthy ways. First, if you have a rare mtDNA haplotype, you may have a not-too-distant relationship to matches you find in an online database. Second, if you’re clever about it, you can use mtDNA to tackle specific genealogical conundrums involving maternal lines. Not sure which of Great-granddad’s 15 kids belonged to each of his three wives? Locate direct-line maternal descendants of his daughters, have them tested, and use the results to assign the children to the correct mothers. Finally, since mtDNA is far more likely than Y-DNA to survive in degraded remains, it’s usually the only option for solving mysteries such as the identity of an unknown soldier.

Ethnic: Generally, DNA tests geared toward identifying particular ancestries, such as African, American Indian, Cohanim (a Jewish tribe) and Tribes of Britain (for example, Celtic and Anglo-Saxon), are your basic Y-DNA or mtDNA tests with additional analysis that compares your results to those typical of a certain ethnicity. If you opt for the standard analysis, the results probably will supply the data you need to do it on your own — provided you don’t mind a little Web surfing.

It’s worth looking into a few ethnic testing companies that have their own databases of test results specific to certain ethnic groups. African Ancestry’s database, for example, includes 10,000-plus people from Africa, and Trace Genetics has a similar collection of Native American samples. Both companies will compare your results to their data in search of a tribal affiliation — and they’ve found one for roughly 70 percent of their customers. (One company says that many of the “misses” occur because the testee’s paternal line turns out to be European.) As with any Y-DNA or mtDNA test, ethnicity tests apply to a single branch of your family tree, so you’ll have to be strategic about who gets tested. If the alleged American Indian in your family was a woman in your father’s maternal line, Dad can be mtDNA-tested to determine a relationship — but since he didn’t pass that mtDNA strand on to you, being tested yourself wouldn’t help.

Biogeographical: This test, also referred to as DNAPrint and ANCESTRY by DNA, has generated the most controversy. (Note: The company that administered Ancestry By DNA tests is no longer in business.) It goes beyond Y-DNA and mtDNA to evaluate autosomal markers — that means it looks not only at the top and bottom lines of your pedigree chart, but at all the branches in between.

The results are expressed in percentages of DNA from various geographic origins, broken into four categories: Indo-European, Sub-Saharan African, Native American and East Asian. For example, one of my family members of Eastern European stock tested as 95 percent Indo-European and 5 percent East Asian. When you think of the proverbial “invading hordes from the East,” it makes sense that she’d have a dash of Asian blood.

SNP: The single nucleotide polymorphism test is another Y-chromosome test that’s gaining popularity. Fortunately, scientists had the good sense to shorten its name to SNP (say “snip”). SNPs are a type of mutation that occurs less frequently than STRs (the focus of the Y-DNA test), but there’s a correlation between the two. In fact, your Y-DNA test results can predict your SNP results — you can use the online tool at <> to make such a guess. So why take the SNP test? To be absolutely sure about your deep ancestry along the top line of your pedigree. It won’t sort out your immediate roots dilemmas, but it’s the test to take if you’re curious about, say, whether your paternal line hails from the Eurasian Steppes, Scandinavia or the Middle East.

Who knows? Maybe you’ll find a cousin. Only you can decide whether DNA testing is a good next step, but for a growing number of genealogists, the answer is a definite yes. If you’re dealing with stubborn brick walls, wishing for a way to verify your traditional research, or struggling with your Smith or Jones line, it might be time for you to take the test.
From the February 2005 Family Tree Magazine