The Autosomal Me – Unraveling Minority Admixture

haystack

You’re invited to come along with me on a journey.  It’s an epic event, a journey into the deepest recesses of our cells, into the smallest pieces of our DNA, into the part previously thought to be useless because it’s so tiny.  It’s the journey to find minority admixture.  Minority in this case means small amounts of admixture.  In my personal situation, this means both Native American and African.

People who have larger amounts of admixture don’t necessarily have to do this, although it can still provide useful information.  If your autosomal percentages are uniformly recognized and reported by most or all of the testing companies, meaning over 1%, you probably know which of your relatives contributed your minority heritage and you don’t need to look for that proverbial needle in a haystack.  Not everyone is that fortunate.  I’m not.  I know of Native heritage through my mother’s Acadian ancestors, but the ones in my father’s side have consistently eluded identification.  It’s there, but where?  It’s haystack time for me!

This past year or two, genetic genealogy has been hallmarked by advances in autosomal DNA and the supporting technologies using tools like 23andMe and Family Tree DNA’s Family Finder tests.  In order to figure out how people are related to you, what level of cousins they are, and which genealogical line they come from, we’ve been using independent tools like phasing, where you compare your DNA to that of your parents or other close known relatives to see who gave you which pieces, or segments, of DNA.  Then, when someone matches you on that segment, you can tell which side of yoru family it came from, and sometimes which genealogical line it came from.  This sets the stage for one day being able to have this conversation with someone:

“Hi John, I see that we are 117th cousins and we have a match at location 17,387,426 on chromosome 3.”  Beth

“Hi Beth, why yes, we are indeed cousins, but we’re actually 115th cousins, 11 times removed instead of 117th cousins.  Our match is through Attilla the Hun’s 37th concubine.”  John

Ok, so maybe I’m dreaming a bit…but this conversation is not just a possibility, it’s a certainty 10 years from now, but perhaps with less dramatic cousinships:)

To date, the rule of thumb for finding ancestors has been that small matches should be disregarded because they are probably identical by state (IBS), not identical by descent (IBD), meaning not useful genealogically.

What’s IBS you ask?  It’s a segment that is typically too small to be counted as an IBD, or identical by descent, segment.  This means that you and the person you match on this small segment descend from a common population, not necessarily that you share a common ancestor within the past several generations.  Genealgoically relevant segments are recognizable because they are larger.  To understand why and how this works, refer to my article, “Autosomal Results, the Basics.”

There is no absolute line in the sand, but generally segments smaller than 7cM (centimorgans) or 700 SNPs (some say 5cM and 500 SNPs) fall into this category.  Dr. Tim Janzen, the genealogical “father of phasing” discards all matches in his spreadsheets less than 3.5cM.  That’s because he’s looking for positive genealogical matches and does not want the data to be cluttered up by possible IBS matches.

However, when you have small amounts of minority ancestry, it stands to reason that these small tidbits could be very useful in identifying which of your genealogical lines produced these small amounts of admixture.  If you can identify which lines provided this minority admixture, then you’re well on the way to identifying which ancestor contributed the minority admixture.

When looking for minority admixture in two related people, finding these small segments in the same location should provide meaningful information and confirm minority heritage.  Said another way, if you both have less than 1% Native heritage, both share a common ancestor, and both carry your less than 1% on the same segment….one might say it’s not likely to be coincidence, especially if there is a pattern across multiple chromosomes/segments.  Identifying the common segments of your common ancestor can lead to identifying the specific family line, especially if you match others as well.  In essence, this is the genetic equivalent of “surround and conquer.”

Let me give a very short example here.

Let’s say I match my mother on part of chromosome 1 that is Native.

Then let’s say I match her first cousin (my first cousin once removed) on mother’s mother’s side on a smaller piece of that same segment.  This immediately tells me that this particular bit of Native heritage is not from mother’s father’s side.

autosomal Hill

Another match to a more distant Hill cousin further defines the path of Native ancestry,  showing that the Native heritage came through mother’s grandfather’s mother’s line.  You can see how we track this ancestry and whittle down the possible sources.

So, I’ve set out to test this minority ancestry tracking theory.  Because we are dealing with such small segments of DNA, “rooting around in the weeds,” as Bennett Greenspan so aptly put it, and have no mechanized tools, this journey is long, tiresome and tedious.  It’s also thrilling.

As with all experiments, I have wondered many times if I was wasting my time.  I’ve completed steps and then redone them a different way when I realized there was a better or more revealing method.  More than once.  That comes with the territory.

I debated about how to share this new technique.  In the past, I would have published this as an academic paper, but with the delays surrounding the publication of JoGG, and the fact that the last paper through JoGG took 18 months to get out the door, much of this information would be stale by then.  I thought about publishing as an e-book as well, but I finally settled on my blog.  I feel that I can reach more people in a much more timely fashion this way.  I also really like the blog because I can write in a more relaxed fashion than I could in any other venue and it gives you the opportunity to interact as well.

I also don’t know what to call this new methodology.  I have just been referring to it as the weeds method, but that’s not very scientific.  I considered the APM technique – Ancestry Population Mapping.  Sounds too nonspecific.  The PTM – Personal Torture Method – nah – puts people off even if it is true.  MAT – Minority Ancestry Tracking – that’s a possibility but isn’t very specific.  Fragmented Chromosome Mapping, FCM, has possibilities.  So, I’m open to suggestions.  If at the end of this series, it’s still the weeds method….well, then weeds it is.

Over the next few weeks, I’ll be writing about this journey, my discoveries, and sharing techniques with step by step instructions so that you can use the same tools.  Join me for the multi-part series, “The Autosomal Me.”

39 thoughts on “The Autosomal Me – Unraveling Minority Admixture

  1. Just wanted to say that I really, really appreciate your blogs. They are so informative and are helping me with understanding the field of DNA as it relates to genealogy. I think FCM is your best choice for an acronym to describe finding minority. Thanks again!

  2. Thank you so much Roberta! I’m currently using my Autosomal test in conjunction with a male cousins YDNA to find the family of our great grandfather who had an assumed surname.  Until DNA, this was literally impossible, but now!  Oh the possibilities! Carla

  3. I think “weeds” is a fun name, but I have a small farm so weeds have a special meaning to me! You could also call it the “Haystack” method, as finding a needle in…..but professionally/scientifically, I like MAT as it is the “minority ancestry” pieces that are the primary focus.

  4. Roberta, I also love to play with these little segments, finding patterns that you cannot explain so easily. If you’re open enough – and at the same time with discipline – these tiny segments start to whisper old stories, old connections.

  5. Excellent article as always. One question that I’ve wondered about for awhile that others may have had who have done autosomal testing –

    My 23andMe results on Ancestry Composition, Chromosomal View, show a segment of Native American, reported to comprise .6% of the total, that’s point 6 percent, so it’s small, but it appears to be concentrated on chromosome 15. The results do not give centimorgan lengths, but do show the segment to be on the bottom part of chromosome 15.

    Using Gedmatch from uploaded FTDNA results, its Ad-Mix Proportions Analysis, Chromosome Painting using Dienekes’ World9 tool, it presents a similar segment on chromosome 15 but appears to highlight from the top.

    My question, finally, is does the presentation of the chromosome segment, whether bottom or top, reflect whether the result is from the paternal line or the maternal line, or is it arbitrary how the testing company shows it? That alone may help others in determining the ancestral line. Thanks Roberta

    • It’s arbitrary unless at 23andMe you have a parent on record and they have phased your data against the parents. As far as the the other tools, entirely arbitrary. We’ll be looking at comparisons and how to tell which side they came from in future blogs in this series.

  6. Pingback: The Autosomal Me – The Ancestors Speak | DNAeXplained – Genetic Genealogy

  7. Pingback: The Autosomal Me – The Ancestors Speak | Native Heritage Project

  8. Dear Roberta,
    I thought I would mention that the techniques that you’re describing about how you are using biogeographical analysis data for purposes of chromosome mapping have been used by others for quite some time. Thus, this is not really a “new technique”. Several years ago I used the Ancestry Painting information at 23andMe to map the portions of my mother’s chromosomes that are Native American in origin. As you know, Ancestry Painting could differentiate between 3 different ancestral populations: European, African, and Asian. Since my mother does not have any Asian ancestry other than her Native American ancestry I was able to attribute her Asian segments that appeared in Ancestry Painting to my mother’s Native American ancestor, Myeerah (b. 1757). I noted other people were doing the same thing with their 23andMe data as reflected in Ancestry Painting. When I sent my mother’s 23andMe data to Doug McDonald a year or so ago he was able to find one additional Native American segment that 23andMe did not report in Ancestry Painting.

    After 23andMe released its Ancestry Composition feature in early December 2012 I also began to map other regions of my mothers’ chromosomes based on the Ancestry Composition data. For instance, I labeled some of her segments as being Iberian in origin, even though I am not sure exactly which ancestor traces his or her ancestry to Iberia at this point in time. I discussed this in a recent post on the RootsWeb autosomal DNA list at http://archiver.rootsweb.ancestry.com/th/read/AUTOSOMAL-DNA/2013-02/1359865230. I was also been able to map my father’s Ashkenazi Jewish DNA segments using 23andMe’s Ancestry Finder data a year or so ago and more recently with Ancestry Composition as well. The start and stop positions of each segment that 23andMe has attributed to a specific population or region in Ancestry Composition may be found using a technique described by CeCe Moore at https://www.23andme.com/you/community/thread/16913/. I am not sure that anyone has given a formal name to the technique we are using, but a reasonable description for this technique would chromosome mapping using biogeographical analysis data.

    Thank you for sharing with her readers your personal information about how you are also using BGA data to map your Native American DNA segments. This entire topic is definitely one that needs much wider discussion than it has had in the past, particularly now that Ancestry Composition and some other very good BGA tools are available.
    Sincerely,
    Tim Janzen

    • Hi Tim,

      I don’t think all of these tools have been used together before to ferrett out specific minority ancestry. If so, it hasn’t been documented as such in a way that people could do for themselves. The new download tool by Rob and at the admix tools together at GedMatch make such a big difference, as do the techniques that I’m using to group clusters of types of segmentation and then applying them separately to the downloaded file of matches and their respective segments. Also, we’ll be evaluating a couple of different approaches and evaluating which one works best in various circumstances. It’s very helpful to be able to compare this technique and these results to my actual pedigree chart where some known Native admixture is found. My goal with this is to document the process, step by step, with illustrations and graphs how this can be done so that those searching for those very small minority admixure pieces can do this for themselves. As you know, I’ve been working on this since the Nat Geo results came back, which was the final vendor piece of the pie. (Tim saw an early draft of this paper a month or so ago.) None of this would be possible without those who have contributed pieces of this pie, as several different tools are being used together, most of which have been developed by other genetic genealogists.

      Roberta

      • Dear Roberta,
        I agree that the admixture tools at GEDmatch have been helpful for chromosome mapping purposes. I think that they are a nice complement to 23andMe’s Ancestry Composition, which I consider to be the current gold standard in this arena, at least in terms of looking at relatively recent ancestry. The nice thing about Ancestry Composition is that you can get the precise start and stop positions for segments that 23andMe has attributed to a specific population. It is harder go get the precise start and stop positions from the Chromosome Painting feature at GEDmatch for the various admixture tools. You have to estimate those start and stop positions. In any case, I have identified additional small Native American segments in my mom’s DNA using the Chromosome Painting feature at GEDmatch for the MDLP Project, Eurogenes, Dodecad, and Harrapa World utilities that don’t appear in Ancestry Composition. Thus, these admixture utilities are indeed helpful as a supplement to Ancestry Composition.

        One major disadvantage of the admixture utilities at GEDmatch is that they can’t currently incorporate the data from closely related relatives (such as in a two parent/one child trio) into the admixture results. This unfortunately limits the specificity of the data you get from these utilities. An important aspect of Ancestry Composition is that when the parents are linked to a child’s results in a family tree the overall accuracy of the results is improved at least to a certain extent over and above what the accuracy would be if the parents weren’t linked to the child.

        Another admixture tool that can be used is the Ancestral Origins feature at deCODEme. Unfortunately, that tool only breaks the data down into three populations: European, African, and East Asian, so it isn’t all that much more informative than 23andMe’s Ancestry Painting was.

        The Geno 2.0 results are interesting, but since Geno 2.0 doesn’t give you a chromosome by chromosome view of your genetic makeup like Ancestry Composition and the admixture tools at GEDmatch do when using the Chromosome Painting feature, I don’t consider the Geno 2.0 results at all helpful in terms of doing chromosome mapping. I am looking forward to reading your additional blogs on this topic.
        Sincerely,
        Tim Janzen

  9. Pingback: The Autosomal Me – Who Am I? | DNAeXplained – Genetic Genealogy

  10. Pingback: The Autosomal Me – Who Am I??? | Native Heritage Project

  11. Roberta, I joined the site to just read and learn about dna results. I had a male cousin do a ydna from Family Tree DNA. For 40 years I have been researching Willburn’s from Virginia to Missouri to Dallas, Texas. Much to my surprise when the results came back to a Urias Martin. I went back and looked in court records; Narcissa Willburn was an immaculate note keeper, to find that her daughter Louisa Willburn married a William Hardin Martin, who was the son of Urias Martin. All this to say, I still can’t find death on Louisa or birth records on my great-grandfather; nor any other information to help definitely tie into this Martin family. By stories told and three of us cousins have been told the same story, I tend to believe it is true. I can trace my great-grandfather from 1872 when he married until he died in 1933. What test would be most beneificial for me to have done now? I have so enjoyed reading all the emails and post. This is all new to me and it is like learning Greek. I just don’t understand all these numbers. Thanks so much for your time and I look forward to your reply. Carroll Vanscoy Robison of Texas

    • Hi Caroll,

      I’m not sure which numbers you’re referring to, but the matches are the important thing. Also, not sure what story you’re referring to either. However, since you’ve taken the Y test already, the other one what would make sense if the Family Finder test. This one matches you with cousins from all your lines. But then of course it’s up to you to figure out how you match.

      Roberta

      • Roberta, Thanks so much for your reply. We took the 37 marker test and we match to Urias Martin in 34 of these markers. Is that close? As soon as possible I will do the Family Finder to see what we get from that. The story I was referring to was about my great-grandfather Wilburn. I have many family lines that come out of Cumberland Gap and will be keeping up with the blog. Is there a really good book that explains in laymen terms dna and its meanings? Have a great day and thanks for your reply again. Carroll

      • There really isn’t an up-to-date book about DNA. One of the reasons I started the blog was exactly that. I’ve tried to hit most of the basics in the various postings. I do understand there are a couple new books in the works though so looking forward to seeing them, hopefully soon.

    • Native Americans were enslaved even before African Americans on this continent and throughout the 1600s and 1700s. There is a lot of Native American ancestry in the slave culture. By the time the slaves were freed, in 1865, they didn’t know what African tribe their ancestors were from nor that they had Native Ancestry. A lot of African American people do have Native Ancestry.

      • Thanks for setting the record straight Roberta. Too many folk think we are/were making these claims based on baseless facts. I have Native Ancestry so says 23andMe.

  12. Pingback: The Autosomal Me – Testing Company Results | DNAeXplained – Genetic Genealogy

  13. Pingback: The Autosomal Me – Rooting Around in the Weeds Using Third Party Tools | DNAeXplained – Genetic Genealogy

  14. Pingback: The Autosomal Me – DNA Analysis – Splitting Up | DNAeXplained – Genetic Genealogy

  15. Pingback: The Autosomal Me – Start, Stop, Go – Identifying Native Chromosome Segments | DNAeXplained - Genetic Genealogy

  16. I am entirely new at this, to say the least. I feel like i’e begun a rocket scientist course with no qualifications what-so-ever.
    I asked Family Tree DNA to see if they could find any Native American in my MTDNA . They say it’s strictly European. Of course, The Y chromosome was not tested because none of my brothers have been tested. My father has been deceased for a long time. My mother, for 12 years. I am looking for a needle in a haystack. Can you help me?

  17. Pingback: The Autosomal Me – Extracting Data Segments and Clustering | DNAeXplained – Genetic Genealogy

  18. Pingback: The Autosomal Me – The Holy Grail – Identifying Native Genealogy Lines | DNAeXplained – Genetic Genealogy

  19. Thank you so much, for this site.
    I am , 0.5 Native,
    and 0.2 African, but can’t find out what side. Not sure what this means.

      • Thanks,
        I guess it is lost in time, My mom ,and dad are both gone. But, I like to think, that my Black, and Native , 6th,or, 7th, grandparents are calling out to me, that they are here in me.

      • If your parents have siblings, you can test them. If you can test several siblings on both sides, that will provide a good “proxy” for your parents.

  20. Pingback: The Autosomal Me – Summary and PDF File | DNAeXplained – Genetic Genealogy

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s