It took a long time for me to dip my toes into the DNA testing waters, being both an avid genealogist and a very private person. But just as giving birth changed my relationship to modesty, starting a blog changed my relationship to privacy. I'm still both modest and private, but not in the same way. The biggest obstacle to DNA testing was knowing I was dragging my family along. As recent events have shown, criminal behavior (and other indiscretions) can be found out by DNA through relatives' information available on genealogy websites.
But I discovered long ago that privacy as we knew it is dead. I remember working with a family researcher who was writing a book on one side of our family. At one time, I would have refused to contribute any information, but had since been helped so much in my research by a book on the Wightman Family that I wanted to help others the same way. The Wightman book, incidentally, has information on me and our family that was contributed without my knowledge or consent. At the time I was not happy, but I got over that and now appreciate it. Except for where the data is wrong....
The point, however, is that while such direct contributions help researchers, they're not all that necessary. When one of my family members declined to contribute his family's information to the project I was helping with, the researcher understood his reluctance—but he added, "Let me show you the information I've already obtained from public sources." He already had just about everything he could use. As Illya Kuryakin Dr. Mallard said on NCIS last night, The Internet will be the death of us. Or at least of privacy.
In light of all this, Porter and I each decided to submit a sample to AncestryDNA.com, and eagerly awaited the results. Later we uploaded the DNA data to MyHeritage.com, and eventually gave another sample to 23andMe.com—the latter for both the ancestry and the health screening.
This post is not for a detailed analysis of the results, but an overall impression of the value of the DNA testing. First, from the point of view of genealogy.
For us, the Ancestry.com screening was the most useful. This is for two reasons.
- They have the largest database from which to work, and that is what makes the testing useful—comparing your DNA to that of other populations. For this reason it is also most useful for those of European background, because of the large numbers of that population who have participated. The testing services are working to improve the experience for under-represented populations, but for now the data is not so robust.
- I have uploaded our family tree, with its nearly 15,000 individuals, to Ancestry.com, and that's largely what makes their DNA service helpful for genealogy. This gives context to our DNA matches, and I've already confirmed known relatives while learning of several more. My tree is at the moment private on Ancestry, which means people have to ask me about the information, which is a good way to get to meet them. Someday I will make it, or at least a version of it, public, but the tree itself isn't ready for that exposure yet.
No doubt MyHeritage would be more useful if I put a tree up there as well, but that's on the "Someday/Maybe" list. I only uploaded our data because at the time they gave free access to their resources if you did. So far they've only found us "third-to-fifth cousins"—tons of them—which is not of much use without trees to compare, and most people seem to have no trees or very small ones. Third cousins share a great-great-grandfather, so it requires a significant amount of family history knowledge to make the connection.
23andMe is in the same situation as far as genealogy goes. So far nothing found even as close as second cousin (sharing a great-grandfather).
How has this helped my genealogy research? Well, through Ancestry.com I've connected with a few previously unknown cousins, a couple close enough to be useful in sharing information. Even the ones that are more distant have been useful in providing some confirmation of my research. Overall I'm glad I took the plunge, if only for this reason. It also has a lot of potential for more and better information as time goes on. One important caveat: There is a lot of error in online family trees. Even with DNA support, this information is best taken as inspiration for further research, and for mutual sharing of data sources.
Now for what most people want out of DNA testing: heritage and ethnicity information. This is an estimate only, and each company has its own data and algorithm for making its "best guess." Sometime after we had our samples analyzed, Ancestry.com upgraded their system and re-analyzed our data. The results were not terribly much different from the first attempt, though probably more accurate.
The analysis from MyHeritage was closer to Ancestry's original analysis. That from 23andMe was different from any of the others, though quite similar overall.
My impression? The DNA analysis is very good as an overall picture, not so good on the details. For example, Porter's great-grandparents came to the United States from Sweden, and it is well known where they lived before emigrating. In fact, when his dad visited Sweden, he was told he looks just like people who live in that area. Thus when his father's AncestryDNA analysis came back showing his largest ethnicity to be Norwegian, we were taken aback. However, the area he's from may be called Sweden, but it's right on the border with Norway. One can definitely say from his DNA that he is of Scandinavian origin, but that he is specifically Swedish comes from genealogy. One must also remember that the smaller percentages are suspect: of the three analyses, 23andMe was the only one that gave me "broadly East Asian and Native American" ancestry, and that was at just 0.1%, so highly doubtful.
Finally, there's the analysis of genetic health data. This comes primarily from 23andMe, though we also paid an extra $10 post facto for Ancestry's "Traits" screening. I've written about the latter experience before. 23andMe analyzes many more traits than Ancestry's small sample, from "Leigh Syndrome, French Canadian Type" carrier status, to estimated risk for late-onset Alzheimer's Disease, to Lactose Intolerance, to Asparagus Odor Detection.
My thoughts? Interesting, but not quite ready for prime time. Where I have independent data it sometimes confirms, sometimes contradicts the DNA reports. Ancestry says I likely have a "unibrow" but 23andMe says the opposite. Both of them say I probably hate cilantro, and I love it. And so on. So I'm taking the rest of what they say with a few grains of salt. I'm sure there's something to it, and that the data will get better with time, but for now it is more entertainment than useful information. Actually, I take that back: Just as DNA ancestry data is useful as a starting point for further research, the discovery of certain traits might be useful for suggesting further, medical, genetic testing.
There's a lot more to DNA analysis for the serious genealogy researcher to investigate, such as sites that will take your data and give you tools to learn much more about which particular genes you and a DNA match share. I'm not there yet; I have too much to do with my regular research to explore that path further. But it, and my data, are there when I'm ready.
Am I glad I decided to "spit in the tube"? Absolutely; I'd do it again and may later go further with it. I'm very grateful to family members who have taken the plunge as well, because that provides a look at the puzzle from more angles. But it's always important not to expect too much. It's never as simple as trading your kilt for lederhosen, as the Ancestry.com ad blithely shows. Plus there's a risk of finding out things you don't want to know—about family or about health. It's a very personal decision and I understand those who are reluctant to take the risk.
