Genetic Genealogy Idea: Tagging DNA!

In the past few weeks, I have spent considerable time contacting DNA matches in 23andMe as first steps towards looking for verifiable connections.  As I go through this process, I know there must be an easier way to leverage all the work we as individual 23andMe users are doing on the site.  For my day job I deal in knowledge management and let me tell ya – there is a lot of knowledge yet to be managed in 23andMe and it gave me an idea.

Currently, when 23andMe has identified someone with whom I share a DNA segment with, time is spent between myself and my match to share our family tree information in the search for shared ancestry.  Sometimes, it may be possible to verify the exact relationship and other times it may not pan out.  Many factors must be considered when evaluating the shared DNA and for the lucky ones, those relationships can be determined.  For example, Randy Majors describes an inspiring success story in verifying a common ancestral couple (his 6th  great-grandparents) with a 23andMe match.  For the sake of argument, let’s pretend that I too were descended from his 6th great-grandparents and had the same DNA segment match as he and his distant cousin.  As the process now stands, despite the fact that Randy and his distant cousin KNOW their shared DNA segment came from that couple, the only way I would be able to find out would be to contact Randy & the distant cousin (assuming they both showed up as matches for me) and go through the long and tedious process of comparing information and/or doing triangulation.

This is where my idea comes in:

  • what if after Randy & the distant cousin confirm this couple as the origin of their shared DNA they were able to “tag” that DNA segment with information and metadata about the couple and add that information to the 23andMe website so that it could be USED BY OTHERS.
  • what if everyone who confirmed relationships could tag DNA segments to specific individuals — imagine how genea-rich with information 23andMe would become!
  • what if when I logged into 23andMe I were presented with names/dates/locations of specific people based on computer aggregation and analysis of all the “tags” that successful matches had been able to identify and associate with DNA segments.
Thus, for the next person who matches Randy at that segment, 23andMe would be able to show them the match to Randy at that location, but also be able show information about the people from which Randy inherited the DNA.  Would that not be awesome!
Of course there are issues to be considered but ultimately I do think genetic genealogy needs to head in this direction and I would like for it to happen sooner rather than later.  Let’s leverage the knowledge of the 23andMe user database and combine it with computing power to facilitate the tagging and presentation to others. They already do it for health traits and conditions so why not for genealogy?
A natural question that comes to my mind is who would do this first? 23andMe has an obvious health focus and I wonder how committed they are to the genealogy aspect of DNA testing. But, is now getting into the autosomal DNA testing game and already has the technology to allow users to tag  specific individuals across their massive database of records.  Imagine them making it happen with DNA.  For this reason, I know that as soon as I can, I will be making sure I get my DNA info in the Ancestry database also.   Think of the potential!!!

What do you think of this idea?






Comments (8)

  1. Leah

    That is a fantastic idea. I’d love to see it happen, if at all possible. I think it could be an especially valuable resource to lineage societies, and applicants to them that are struggling with the needed documentation but have the genetic proof.

  2. taneya (Post author)

    Hi Leah – glad you like the idea. I hope one of these companies is able to make it happen! And yes, it would be helpful for lineage societies. Thanks for sharing your comment.

  3. Ginger Smith

    Hi Taneya, these are great ideas and would probably work well. The only issue I can foresee is running into the same problem with online family trees where people mistakenly tag someone to a segment or they tag them because they “think” it should be tied to an ancestor.

    Now the other issue that with ftDNA there are a lot more shared segments. I noticed while looking at my friend’s 23andMe results that there are only 1 or 2 segments that are shared, whereas with ftDNA there are on average about 6-10 shared segments. It is often hard to determine which segment belongs to which name.

    But I do something similar to this. Once I find a known connection, I have that match as a “gold standard” and anyone who also matches to them I try to work with first and I use my “gold standards” as guides.

  4. taneya (Post author)

    Thanks Ginger! Yes, the issue you raise of people “thinking” it comes from a couple is one I too thought about. But, as we know – even traditional genealogy records like that have to be verified, but maybe they could offer some clues :-). With regard to ftDNA, I was not aware of that since I have never used ftDNA. Interesting, and I agree it would make it more complicated. Perhaps algorithms could be used to “cluster” results and again, though it may not be perfect, maybe it could be a guide. I am really looking forward to seeing how genetic genealogy develops over the next decade.

  5. Robert


    Thanks for the tweet yesterday on this topic. I see that you have been thinking about a tagging concept for DNA segments. There is a lot that can be done in this area. One of the basic things I have found is to sort the Excel Ancestry Finder file from 23andme. I have found some hard and soft matches, then extended genome sharing invitations, and tried to find genealogical confirmation. I have also used to confirm matches and ancestry from 23andme. My knee jerk reaction from the panel at Rootstech is that expecting to isolate matches from a certain village with a time date is too much to expect. On second thought, we need to reach for the sky when trying to extend the technology of the human genome.

  6. taneya (Post author)

    Hi Robert – all the approaches you mention are ones I’ve been doing and the process has been extremely helpful. I do think we need to shoot for the stars! I believe it is entirely conceivable that DNA can be used to isolate to specific geographic regions. In my own case, I have DNA segments that I can attribute to specific individuals and that couple lived in the same town their entire life. If we as users can participate and add value, I know we will see this vision come true one day. I can’t wait!

  7. Kari Lemons

    I just started a Autosumnal testing project at FTDNA for all the children adopted from Cambodia to the US. My boys were the first tested for the project. What I have learned is that FTDNA does not break out the test results for mother and father. The results are combined. To get a break out ofparental results after the family finder test, they require a male to get a 67-111 marker Y-DNA test and a female to get the MTDNA test. Then they will calculate which results are the father or mother.

    It probably ends up costing the same as the [email protected] test does, to get the same results with one test. I think they do it because they know a person is more likely to spend money on a less expensive test but then get facinated and want more info and spend for another test.

  8. taneya (Post author)

    thx for the comment Kari. I have learned a lot since doing genetic genealogy and know there is so much more to learn. I’m just now getting familiar with FTDNA results.

Comments are closed.