First looks at AncestryDNA

First looks

The new kid on the autosomal DNA block is AncestryDNA from Ancestry.com, joining the Family Finder test from Family Tree DNA and the Relative Finder test from 23andMe.

Autosomal DNA testing, remember, is the kind that works across gender lines so you don’t have to find a direct male line from father to son to son (YDNA or Y-DNA1) or a direct female line from mother to daughter to daughter (mitochondrial DNA or mtDNA.2) It tests DNA from all of the chromosomes except the gender-linked X and Y chromosomes,3 and helps you identify cousins to share research with.

And when this self-confessed DNA junkie was handed a test kit at the National Genealogical Society conference,4 well, hey… I’ve said repeatedly I’ve never met a DNA test I wouldn’t take. My results came in this past week, and I’ve had a chance to play with the system enough to have a sense of where it shines and where it needs more work.

Admixture results

The first thing you see when you open your AncestryDNA result page is the admixture information. That’s the analysis of your DNA to try to determine what “genetic percentages” of your ancestry came from particular continents or regions.5

Here are my “genetic percentages”:

This, to me, is one place where AncestryDNA shines. Comparable analysis at 23andMe simply reports me as 100% European. At Family Tree DNA, I get a little more information — it reports me as 85.86% Western European (French, Orcadian, Spanish) and 14.14% European (Tuscan, Finnish, Romanian, Russian) — but both have a large margin of error of 11.65%.

Here, you can see, it looks like AncestryDNA has nailed it: my father’s solidly-German ancestry is clearly reflected in the 43% Central Europe and the 7% Finnish/Volga-Ural, and my mother’s colonial American ancestry shows up in the 35% British Isles and 15% Scandinavian. That last bit is likely to be the result of waves of Vikings and other invaders to Britain.6

Clicking on the See Full Results button takes you to more information, and the opportunity to zoom in on a map to see where the birth locations of people in your family tree may intersect with these “genetic percentages”:

So to the extent that you’re looking for clues to your geographic origins, this test is a strong contender.

The match system

Linking up with cousins to extend your knowledge of your family tree is the big reason to do autosomal DNA testing. Here, AncestryDNA has some outstanding features — and needs a lot of work.

Matches are presented in a chart form with those most closely related genetically shown first:

Clicking on the Review Results button takes you to the details of your match — and it’s here that AncestryDNA shows its major strength — and its major weakness — in one and the same fature. It’s the fact that the results are tightly integrated with Ancestry’s online family tree system.

What you get presented with, first, is a quick comparison of your “genetic percentages” with the “genetic percentages” of the match:

Without the specific underlying genetic information, however — how much DNA you share with the other person and where in your DNA (on what chromosome or chromosomes) the common areas are — this comparison is less than useful. That level of detail is something other testing companies do offer, and that AncestryDNA may eventually offer down the road.

Under that you see your match’s family tree, and if someone in your tree intersects with someone in your match’s family tree, you’re presented with that information immediately. I’d be happy to show you what that looks like, except I don’t have any matches that intersect with my tree anywhere in my top 50 or so matches… Sigh…

Where the trees don’t intersect — and that will be more common particularly until the database gets built up and people work on their family trees — you’re given a list of surnames you share in common, and you see your match’s tree. You can click on any individual surname to compare it to your tree, and you have a very useful option to switch over to a map to see where the birthplaces of folks in your tree may be in geographic proximity to the birthplaces of folks in your match’s tree:

Why this tree-matching is the great strength of the system is clear: if two people have well-researched family trees on Ancestry and a common ancestor appears in both trees, that’ll be highlighted immediately. Being able to see your ancestral origins on a map with the ancestral origins of a match is very helpful since often the real leads with this test are found not with surnames in common but rather when you find two families were in the same place at the same time.

But it’s equally clear why it’s the great weakness: poorly-researched trees will show “common ancestors” that aren’t common at all, and if the other person doesn’t have a tree, the match is essentially useless. There’s no way to contact a match except through the Ancestry system, and — so far — there’s just not enough genetic information independent of the trees to help overcome brick walls.

So my take on this: it’s a mixed bag that’s best for beginners in genetic genealogy. Not a bad way to start off for those who want a quick and easy way to see where they have matches, and attractively priced at the introductory price of $99.

The interface

Once you get past the wealth of information being presented in a very attractive package, the one area where AncestryDNA most clearly does not shine is in the area of the interface. It needs a lot of work, and AncestryDNA is well aware of that. There’s a good reason why this is labeled a Beta system and why it’s described as a work-in-progress.

First and foremost, the system is sloooooooooow. Each page of results takes a long time to load even on my very high speed home cable internet connection — minimum 15-20 seconds, often longer, and it then takes another 8-10 seconds for a match’s results to appear once I click on the Review Results button. When you go back to the match list, it still takes the full 15-20 seconds or more for the page to load. I’m told there are “good improvements” coming on this that are “right around the corner.”

There are only two ways to sort matches: by relationship and by date. And there are only two available filters: a reviewed-vs.-unreviewed filter (whether you’ve already looked at this match); and a starred-vs.-unstarred filter (you can highlight a match by clicking on a star icon to highlight that match for further review).

It’d be terrific to be able to easily find matches where my tree intersects theirs or where we have particular surnames in common — and neither of those options is available yet. There’s no way yet to add a note to a match. And there’s no way as yet to compare two matches to each other, rather than just to me. All of these are features that are reported to be in the works.

There’s no way at the moment for the system to tell you that a match whose results you’ve already reviewed has changed something in his or her online family tree. That change could highlight an ancestor in common or correct an error — and right now, you wouldn’t see it at all. That, too, is something that’s said to be on the horizon.

The underlying data

Right now, AncestryDNA doesn’t make any of the underlying raw genetic data available. John Pereira, vice president of DNA at Ancestry.com, has long said the company understands that access to the raw data is “important to serious genetic genealogists and we’ve got that under review. We’re working to figure it out.”

Without that access, serious genetic genealogists simply aren’t going to make this test a priority. It’s just not useful to have a match described as “distant cousin, moderate confidence” or “distant cousin, very low confidence” without knowing what those terms mean. How much DNA do I share with that match? How long is the longest segment I share?

Conclusion

Genetic genealogy is one of many tools we have today that earlier generations of genealogists couldn’t have dreamed of. Integrating paper-trail family tree genealogy with genetic genealogy — in the long run — is going to take us places neither could do alone.

The theory behind what AncestryDNA is doing is first class, and the science is first class. The interface needs a lot of work to speed it up and make it easier to use, and access to the underlying data should be moved up the priority list.

And the ultimate utility of this whole approach is only going to be as good as the work people put in to creating an accurate database of family trees. In a very real sense, this test using this presentation method is only going to be as good as its users make it.


 
SOURCES

  1. ISOGG Wiki (http://www.isogg.org/wiki), “Y chromosome DNA test,” rev. 23 Jul 2011.
  2. ISOGG Wiki (http://www.isogg.org/wiki), “Mitochondrial DNA,” rev. 30 Jul 2010.
  3. ISOGG Wiki (http://www.isogg.org/wiki), “Autosomal DNA,” rev. 8 Feb 2012.
  4. Truth in blogging: I was given this test kit by Ancestry.com. The fact that I didn’t pay for it has in no way affected my objectivity with respect to its results.
  5. ISOGG Wiki (http://www.isogg.org/wiki), “Genealogical DNA test,” rev. 3 Feb 2012.
  6. For a good understanding of admixture results from this test and the labels applied to various DNA results, see Blaine Bettinger, “Problems with AncestryDNA’s Genetic Ethnicity Prediction?,” posted 19 Jun 2012, The Genetic Genealogist (http://www.thegeneticgenealogist.com : accessed 4 Aug 2012). See also CeCe Moore, “My Review of AncestryDNA’s Admixture Tool and a Glimpse into the Future of Genetic Genealogy,” Your Genetic Genealogist, posted 26 Jun 2012 (http://www.yourgeneticgenealogist.com : accessed 5 Aug 2012).
Print Friendly
This entry was posted in DNA. Bookmark the permalink.

38 Responses to First looks at AncestryDNA

  1. Chris Staats says:

    Thanks for the review, Judy – very timely, given that I just ordered mine earlier in the week.
    But, ugh…so if I want to make any use of matches, it appears I need to create an Ancestry tree? Am I like the one who does not already have a tree on ancestry?
    (Well, there is the one that I submitted 10 or 12 years ago so I could access the site for the free two-weeks, or whatever the trial period was…but let’s never speak of that tree, okay?)

    • Judy G. Russell says:

      Yep, you’ve got to put together a family tree to make it useful. Anybody who doesn’t do that is going to get just about no use from this test whatsoever. And you’ve got about four weeks to get it done — that’s how long it’ll take from submitting the test to getting your results.

  2. Nan Harvey says:

    I’ve had about 10 pop up with our common ancestor listed. The most exciting was a fairly close match through my maternal grandmother as that is the first I’ve verified through her line. I wish I knew what chromosome it was on!

    I’ve also had two that showed our common ancestor as being in a line that I have been trying to verify. Since it’s colonial and pretty far back I didn’t jump to conclusions, though, and in both of those trees I found another common ancestor on a completely different line. Fortunately both I and each match had trees that went back enough generations to see this.

  3. Nancy Schlegel says:

    Having submitted my test a week after yours, hopefully my results will be showing up this coming week; can’t wait!

    Here’s hoping for similar clarification vs. my Family Tree DNA genetic percentages. And the map-plotting of my vs. a match’s birthplaces looks really useful, especially in comparing places in those German and Austrian areas that changed political boundaries so often!

    • Judy G. Russell says:

      I was very pleased at how fast the test results came in, Nancy, so yours should be coming along soon. And yes, I think the maps have the potential for being very helpful especially once people realize they have to do very good paper-trail tree work for this to be the tool it can be.

  4. Thanks for the review. I’ve been wanting to “get into DNA” but not really sure where to start. Always learn a lot from reading your posts.

  5. John Boggess says:

    I just got my results back last night, so have not had much time to check things out, but was pleased to have two clear matches show up in AncestryDNA. Both were predicted 4th-6th cousins — one a 3rd cousin on my mother’s side and the other a 5th cousin on my father’s side. One I had already found through other sources, but the second match is new to me. Guess what I will be working on later today!

    • Judy G. Russell says:

      That’s great, John! I don’t have any matches closer than 4th cousin, and no really obvious bingo-here’s-the-common-line matches among them. Sigh…

      • John Boggess says:

        As you said earlier, Judy, it will take time but results should improve dramatically as more people participate and submit family trees. There is so much potential here as word spreads that it really does work!

  6. hausa84 says:

    Had my results for two months now. Total of 20 matches – 2 with 96% confidence that have been verified by paper trail/tree, 3 moderate confidence where we are both clueless about who our common ancestor is and the remainder low to very low confidence – 1 of whom has been verified by the tree as 6th cousin.

    Agree with article – at end of day this may be most powerful tool out there, but at this point in development it is “nice”…

    • Judy G. Russell says:

      If this develops the way it could develop, it has the potential to be the best thing for genealogists since sliced bread. But that is a big “if”… And as to that clueless feeling, I know it only too well. Sigh…

  7. Diana says:

    Thanks so much for this post. I received my kit yesterday! Great information and happy to know that I have four weeks to get my trees up at Ancestry.

  8. Susan Johnson says:

    Send my kit in last week, so maybe, I’ll get the results in about 3 weeks. Thanks so much for this post.

  9. Diane Harman-Hoog says:

    This makes it less than useful for adoptees who do not have family trees. Also in the experience of the members of the AdoptionDNA yahoogroup, the ethnic prediction is mostly wildly off base.

    • Judy G. Russell says:

      It absolutely is less than useful for anybody who doesn’t have a family tree to compare. That includes folks who just haven’t put one together, as well as adoptees and those who discover a non-paternity event in their backgrounds. As for the ethnic prediction, I can only report what I see. I’m aware of the criticisms, and also aware of the caveats on the website that nothing is etched in stone. It’ll be very interesting to see how I test when the Geno 2.0 project kits are available.

  10. Thank so much for the info, Judy. I am waiting for my results. I don’t have a tree on Ancestry. I think I will GEDCOM just my direct line ancestors to narrow things down a bit. I still want to do a yDNA but I have to talk an uncle into it first. Neither of my living uncles have any interest.

  11. George Jones says:

    Judy,

    Thanks for your review. I agree with what you said here:

    “Right now, AncestryDNA doesn’t make any of the underlying raw genetic data available. John Pereira, vice president of DNA at Ancestry.com, has long said the company understands that access to the raw data is “important to serious genetic genealogists and we’ve got that under review. We’re working to figure it out.”

    Contrast the Pereira statement to that of Spencer Wells in regards to the new Geno 2.0 DNA test:

    ** “Your data belongs to you,” he said. “We feel that this is a cornerstone of ethical DTC genetic testing.”

    Hence, I seriously question the DTC Consumerist values of Ancestry.com as a company.

    IMHO, they are failing consumers by not releasing data to consumers RIGHT NOW!

    For sure, it appears they will / or have released data (anonymous / aggregated … I hope) to pharmaceutical and medical related companies and make a buck on the backside by selling that information to such concerns. There is currently a book on the NY Times Best Seller Book List how large organizations tend to unethically treat persons in regards to their private medical data. It them and it will probably happen on DNA Data that can be used years and years into the future. http://www.nytimes.com/2010/02/07/books/review/Margonelli-t.html

    Also, Geno 2.0 has pledged to not collect data on medically reliant SNPs. “The resulting chip includes approximately 146,000 SNPs, avoiding all known medically relevant markers and exclusively concentrating on ancestry informative ones.” Has Ancestry.com taken this same pledge? I don’t think so and perhaps that is the primary reason they are hesitant to release data from DNA Tests which belongs to the consumer!

    Other than GedMatch, what other 3rd parties are allowing comparison of Autosomal DNA Data at >1cM Segments for Consumers between data collected from FTDNA, 23andMe, and potentially Ancestry.com?

    How many of you are using simple MS Excel Spreadsheets?

    How many of you are engaged in the exploration for Rare Autosomal SNPs which may help better to ID putative relatives / putative genetic linkages? I am currently engaged in such an exploration for persons with a Welsh Heritage with connections to the R1b-L371 Y-DNA Haplogroup. I examine matching Autosomal segments between 1 – 3 cMs searching for very distant cousin relationships originating in Wales prior to 1700. This research leads me to fine scaled Genetic Homeland regions in Wales. This research is similar to the People of the British Isles Study.

  12. George Jones says:

    As a followup, this may help to further explain my above perspective / comments.

    In the AncestryDNA Terms and Conditions there are some keywords the management and their attorneys most likely selected for their intent as a “Legal or Contractual Out” for AncestryDNA to use RIGHT NOW the DNA results of a group (1 or more persons) any way they wish. One keyword is INDIVIDUAL (1 person) as noted below in both their old and new versions. “… and not for INDIVIDUAL medical or diagnostic purposes.” So if AncestryDNA used your DNA results and my DNA results (a GROUP of 2) it appears they have the legal right (perhaps not a clearly defined ethical right)they can do anything they wish with it for monetary gain such as putting a patent on it or some other device of ownership. I do not consider that INFORMED CONSENT but rather DUBIOUSLY OBTAINED CONSENT http://dna.ancestry.com/legal/termsAndConditions.aspx

    So, as a lay person, and not as a law person, I have questions concerning what AncestryDNA can do with an Individual’s DNA Data or a Group’s DNA data.

    Industry trade groups, ISOGG and other Genealogy organizations have little or no influence on issues such as this and eventually I see the need for Federal Rules or Federal regulation in this area for DTC DNA tests for Genetic Genealogy purposes.

    • Judy G. Russell says:

      George, I think you’re taking that comment out of context. The language you quote is preceded by this limiting language: All DNA testing performed by AncestryDNA on samples submitted for testing or by uploading a digital version of a DNA analysis is done for genealogical research only, including population and ethnic group-related analyses.

  13. A J says:

    With regard to the release of raw DNA to customers, John Pereira, vice president of DNA at ancestry.com said ” We’re working to figure it out.” Geno 2.0 is $199 versus $99 for the autosmal ancestryDNA test . Geno 2.0 tests yDNA, mtDNA & autosomal DNA and they allow you to download your data. No Brainer! Go with Geno 2.0 unless ancestryDNA “figures it out.”

    • Judy G. Russell says:

      Apples and oranges here. As much as I love Geno 2.0 — and I can’t wait — the fact is, it isn’t going to test nearly as much of your autosomal DNA as any of the tests from AncestryDNA, Family Tree DNA or 23andMe. It’s for deep ancestry, not genealogy. If you want to find cousins or use autosomal results for near-term genealogy, you’re going to need to test with one or more of the three commercial companies.

  14. Nancy Schlegel says:

    Now that I have my own Ancestry DNA results (yes!), went back over your post to see if I missed anything… and I did: the maps comparing locations to other’s tree. Thanks for pointing out!

    One thing not mentioned that I found helpful was the tabs per ethnic element. First, each tab includes list of just those matches on the particular ethnic element. Also, as a newbie DNAer, i learned several things from the tab descriptions, e.g. Ashkenazi Jews may have descended from a small Jewish group in Italy over 1000 years ago.

    This same description also had the following useful info:
    “Are you surprised by the number of matches? … Well, there’s a good reason. It’s a little complicated and science-y, but the bottom line is that it appears our system returns inaccurate matches for people of European Jewish descent. The good news is that our match predictions will improve over time as we grow our database of DNA signatures. In the meantime, don’t be afraid to reach out–you may just discover that distant cousin you never knew you had.”

    FYI, don’t think I had one slow moment and I’m running with basic cable speed – so perhaps those “around-the-corner” changes could have been made this past week?

    • Judy G. Russell says:

      It sure does look as though those changes have kicked in. The page navigation is MUCH faster now, and that’s been reported by lots of users!

  15. David says:

    I wanted to ask, is the ancestry info separated by your father’s or mother’s side or are they just kind of lumped together? For example, would it give you a separate list of your genetic mixture on your father’s side and a separate list for your mother’s?

    • Judy G. Russell says:

      They’re lumped together in the autosomal test, David, because all of your autosomal DNA is a random mix of your Dad’s DNA and your Mom’s. To trace just your father’s side, you want to test your YDNA and to test just your mother’s side you want to test your mtDNA. Different tests for different purposes.

  16. B W Clark says:

    Hi,
    It seems as though you answered this question indirectly, but I’m not all that good at reading between the lines. Here’s my question:

    I took the AncestryDNA test and I received numerous potential matches, one a projected 3rd cousin and after contacting that person, they were indeed my 3rd cousin. Other researchers have taken ONLY either the yDNA or the mtDNA and not the AncestryDNA. Are any yDNA or mtDNA test results from those people included in my AncestryDNA potential matches? OR, must those people who took the earlier y and mt tests need to take the new ancestryDNA to be included in my potential matches?

    Great article, thank you.
    BW Clark

    • Judy G. Russell says:

      Thanks for the kind words. Yes, to be included in your AncestryDNA matches, the other person must have taken the new AncestryDNA test. That’s because the matching is done by a totally different system, looking at a totally different part of the DNA, than the earlier yDNA or mtDNA tests.

  17. Voting Ancestry says:

    My vote is with Ancestry. Their databases are just too massive to be beat.

    • Judy G. Russell says:

      Their database size is essentially irrelevant without the availability of the raw data with tools that allow for real comparison of the data. Knowing that I match somebody without any effective way to really investigate how I match that person isn’t a whole lot of help.

  18. Lindsay says:

    I was wondering if you know where my German ancestry might have gone? I took the DNA test and was shown to have these results: 70% British Isles, 15% Scandinavian, 13% Persian/Turkish/Caucasus, and 2% uncertain. My grandfather has a lot of German ancestry, and I was wondering why that didn’t show up?

    • Judy G. Russell says:

      Lindsay, I keep singing the same old song: it ain’t soup yet. All of these numbers are little more that cocktail-party-conversation pieces so far. The science just isn’t there to support the claims yet. Remember that what they’re doing is comparing your DNA to the DNA of people in their database (not people in general) who live in the places today where our ancestors may have lived years ago (so there’s that mismatch as well). And AncestryDNA in particular has already announced it will be revising its admixture percentages later this year particularly because of the overreporting of Scandinavian in all our results. So sit tight — we have literally years of scientific research to go before the numbers reported actually mean anything.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>