Family Tree DNA Releases myOrigins

my origins

On May 6th, Family Tree DNA released myOrigins as a free feature of their Family Finder autosomal DNA test.  This autosomal biogeographic feature was previously called Population Finder.  It has not just been renamed, but entirely reworked.

Currently, 22 population clusters in 7 major geographic groups are utilized to evaluate your biogeographic ethnicity or ancestry as compared to these groups, many of which are quite ancient.

my origins regions

Primary Population Clusters

  • Anatolia & Caucasus
  • Asian Northeast
  • Bering Expansion
  • East Africa Pastoralist
  • East Asian Coastal Islands
  • Eastern Afroasiatic
  • Eurasian Heartland
  • European Coastal Islands
  • European Coastal Plain
  • European Northlands
  • Indian Tectonic
  • Jewish Diaspora
  • Kalahari Basin
  • Niger-Congo Genesis
  • North African Coastlands
  • North Circumpolar
  • North Mediterranean
  • Trans-Ural Peneplain

Blended Population Clusters

  • Coastal Islands & Central Plain
  • Northlands & Coastal Plain
  • North Mediterranean & Coastal Plain
  • Trans-Euro Peneplain & Coastal Plain

Each of these groups has an explanation which can be found here.

Matching

Prior to release, Family Tree DNA sent out a notification about new matching options.  One of the new features is that you will be able to see the matching regions of the people you match – meaning your populations in common.  This powerful feature lets you see matches who are similar which can be extremely useful when searching for minority admixture, for example.  However, some participants don’t want their matches to be able to see their ethnicity, so everyone was given an ‘opt out’ option.  Fortunately, few people have opted out, less than 1%.

Be aware that only your primary matches are shown.  This means that your 4-5th cousins or more distant are not shown as ethnicity matches.

Here’s what the FTDNA notification said:

With myOrigins, you’ll be able compare your ethnicity with your Family Finder matches. If you want to share your ethnic origins with your matches, you don’t need to take any action.  You’ll automatically be able to compare your ethnicity with your matches when myOrigins becomes available.  This is the recommended option. However, we do understand that sharing your ethnicity with your matches is your choice so we’re sending you this reminder in case you want to not take part (opt-out). To opt-out, please follow the instructions below. *

  1. Click this link.
  2. If you are not logged in, do so.
  3. Select the “Do not share my ethnic breakdown with my matches. This will not let me compare my ethnicity with my matches.” radio button.
  4. Click the Save button.

You can get more details about what will be shared here.  You may also join our forums for discussion* You can change your privacy settings at any time. Thus, you may opt-out of or opt back into ethnic sharing at a later date if you change your mind.

What’s New?

Let’s take a look at the My Origins results.  You can see your results by clicking on “My Origins” on the Family Finder tab on your personal page at Family Tree DNA.

Ethnicity and Matches

Your population ethnicity is shown on the main page, as well as up to three shared regions that you share with your matches.  This means that if you share more than 3 regions with these people, the 4th one (or 5th or 6th, etc.) won’t show.  This also means that if your match has an ethnicity you don’t have, that won’t show either.

my origins ethnicity

Above, you see my main results page.  Please note that this map is what is known as a heat map.  This means that the darkest, or hottest, areas are where my highest percentages are found.

Each region has a breakdown that can be seen by clicking on the region bar.  My European region bar population cluster breakdown is shown below along with my ethnicity match to my mother.

my origins euro breakdown

And my Middle Eastern breakdown is shown below.

my origins middle east breakdown

Ethnicity Mapping

A great new feature is the mapping of the maternal and paternal ethnicity of your Family Finder matches, when known.  How does Family Tree DNA know?  The location data entered in the “Matches Map” location field.  Can’t remember if you completed these fields?  It’s easy to take a look and see.  On either the Y DNA or the mtDNA tabs, click on Matches Map and you’ll see your white balloon.  If the white balloon is in the location of your most distant ancestor in your paternal line (for Y) or your matrilineal line for mtDNA (your mother’s mother’s mother’s line on up the tree until you run out of mothers), then you’ve entered the location data and you’re good to go.  If your white balloon is on the equator, click on the tab at the bottom of the map that says “update ancestor’s location” and step through the questions.

ancestor location

If you haven’t completed this information, please do.  It makes the experience much more robust for everyone.

How Does This Tool Work?

my origins paternal matches

The buttons to the far right of the page show the mapped locations of the oldest paternal lines and the oldest matrilineal (mtDNA) lines of your matches.  Direct paternal matches would of course be surname matches, but only to their direct paternal lines. This does not take into account all of their “most distant ancestors,” just the direct paternal ones.  This is the yellow button.

The green button provides the direct maternal matches.

my origins maternal matches

Do not confuse this with your Matches Map for your own paternal (if you’re a male) or mitochondrial matches.  Just to illustrate the difference, here is my own direct maternal full sequence matches map, available on my mtDNA tab.  As you can see, they are very different and convey very different information for you.

my mito match map

Comparisons

By way of comparison, here are my mother’s myOrigins results.

my origins mother

Let’s say I want to see who else matches her from Germany where our most distant mitochondrial DNA ancestor is located.

I can expand the map by scrolling or using the + and – keys, and click on any of the balloons.

my origins individual match

Indeed, here is my balloon, right where it should be, and the 97% European match to my mother pops up right beside my balloon.  The matches are not broken down beyond region.

This is full screen, so just hit the back button or the link in the upper right hand corner that says “back to FTDNA” to return to your personal page.

Walk Through

Family Tree DNA has provided a walk-through of the new features.

Methodology

How did Family Tree DNA come up with these new regional and population cluster matches?

As we know, all of humanity came originally from Africa, and all of humanity that settled outside of Africa came through the Middle East.  People left the Middle East in groups, it would appear, and lived as isolated populations for some time in different parts of the world.  As they did, they developed mutations that are found only in that region, or are found much more frequently in that region as opposed to elsewhere.  Patterns of mutations like this are established, and when one of us matches those patterns, it’s determined that we have ancestry, either recent or perhaps ancient, from that region of the world.

The key to this puzzle is to find enough differentiation to be able to isolate or identify one group from another.  Of course, the groups eventually interbred, at least most of them did, which makes this even more challenging.

Family Tree DNA says in their paper describing the population clusters:

MyOrigins attempts to reduce the wild complexity of your genealogy to the major historical-genetic themes which arc through the life of our species since its emergence 100,000 years ago on the plains of Africa. Each of our 22 clusters describe a vivid and critical color on the palette from which history has drawn the brushstrokes which form the complexity that is your own genome. Though we are all different and distinct, we are also drawn from the same fundamental elements.

The explanatory narratives in myOrigins attempt to shed some detailed light upon each of the threads which we have highlighted in your genetic code. Though the discrete elements are common to all humans, the weight you give to each element is unique to you. Each individual therefore receives a narrative fabric tailored to their own personal history, a story stitched together from bits of DNA.

They have also provided a white paper about their methodology that provides more information.

After reading both of these documents, I much prefer the explanations provided for each cluster in the white paper over the shorter population cluster paper.  The longer version breaks the history down into relevant pieces and describes the earliest history and migrations of the various groups.

I was pleased to see the methodology that they used and that four different reference data bases were utilized.

  • GeneByGene DNA customer database
  • Human Genome Diversity Project
  • International HapMap Project
  • Estonian Biocentre

Given this wealth of resources, I was very surprised to see how few members of some references populations were utilized.

Population N Population N
Armenian 46 Lithuanian 6
Ashkenazi 60 Masai 140
British 39 Mbuti 15
Burmese 8 Moroccan 7
Cambodian 26 Mozabite 24
Danish 13 Norwegian 17
Filipino 20 Pashtun 33
Finnish 49 Polish 35
French 17 Portuguese 25
German 17 Russian 41
Gujarati 31 Saudi 19
Iraqi 12 Scottish 43
Irish 45 Slovakian 12
Italian 30 Spanish 124
Japanese 147 Surui 21
Karitiana 23 Swedish 33
Korean 15 Ukrainian 10
Kuwaiti 14 Yoruba 136

In particular, the areas of France, Germany, Norway, Slovakia, Denmark and the Ukraine appear to be very under-represented, especially given Family Tree DNA’s very heavy European-origin customer base .  I would hope that one of the priorities would be to expand this reference data base substantially.  Furthermore, I don’t see any New World references included here which calls into question Native American ancestry.

Webinar

Family Tree DNA typically provides a webinar for new products as well as general education.  The myOrigins webinar can be found in the archives at this link.  It can be viewed any time.  https://www.familytreedna.com/learn/ftdna/webinars/

Accuracy

How did they do?  Certainly, Family Tree DNA has a great new interface with wonderful new maps and comparison features.  Let’s take a look at accuracy and see if everything makes sense.

I am fortunate to have the DNA of one of my parents, my mother.  In the chart below, I’m comparing that result and inferring my father’s results by subtracting mine from my mother’s.  This may not be entirely accurate, because this presumes I received the full amount of that ethnicity from my mother, and that is probably not accurate – but – it’s the best I can do under the circumstances.  It’s safe to say that my father has a minimum of this amount of that particular population category and may have more.

Region Me Mom Dad Inferred Minimum
European Coastal Plain 68 17 51
European Northlands 12 7 5
Trans Ural Peneplain 11 10 1
European Coastal Islands 7 34 0
Anatolia and Caucus 3 0 3
North Mediterranean 0 34 0
Circumpolar 0 1 0
Undetermined* 0 0 40

*The Undetermined category is not from Family Tree DNA, but is the percentage of my father not accounted for by inference.  This 40% is DNA that I did not inherit if it falls into a different category.

Based on these results alone, I have the following observations.

    1. I find it odd that my mother has 34% North Mediterranean and I have none. We have no known ancestry from this region.
    2. My mother does have one distant line of Turkish DNA via France. I have presumed that my Middle Eastern (now Anatolia and Caucus) was through that line, but these results suggest otherwise.
    3. My mother’s Circumpolar may be Native American. She does have proven Native lines (Micmac) through the Acadian families.
    4. These results have missed both my Native lines (through both parents) and my African admixture although both are small percentages.
    5. The European Coastal Plain is one of the groups that covers nearly all of Europe. Given that my mother is 3/4th Dutch/German, with the balance being Acadian, Native and English, one would expect her to have significantly more, especially given my high percentage.
    6. The European Coastal Island percentages are very different for me and my mother, with me carrying much less than my mother.  This is curious, because she is 3/4th German/Dutch with between 1/8th and 3/16th English while my father’s lines are heavily UK.  My father’s ancestry may well be reflected in European Coastal Plain which covers a great deal of territory.

What We Need to Remember

All of the biogeographic tools, from Family Tree DNA, 23andMe and Ancestry, are “estimates” and each of the tools from the three major vendors rend different results.  Each one is using different combinations of reference populations, so this really isn’t surprising.  Hopefully, as the various companies increase their population references and the size of their reference data bases, the results will increasingly mesh from company to company.  These results are only as good as the back end tools and the DNA that you randomly inherited from your ancestors.

Furthermore, we all carry far more similar DNA than different DNA, so it’s extremely difficult to make judgment calls based on ranges.  Europe, for example, is extremely admixed and the US is moreso.  The British Isles were a destination location for many groups over thousands of years.  Some of the DNA being picked up by these tests may indeed be very ancient and may cause us to wonder where it came from.  In future test versions, this may be more perfectly refined.

There is no way to gauge “ancient” DNA, like from the Middle East Diaspora, from more contemporary DNA, only a thousand years or so old, once it’s in very small segments.  In other words, it’s all very individual and personal and pretty much cast in warm jello.  We’ve come a long way, but we aren’t “there” yet.  However, without these tools and the vendors working to make them better, we’ll never get “there,” so keep that in mind.

While this makes great conversation today, and there is no question about accuracy in terms of majority ancestry/ethnicity, no one should make any sweeping conclusions based on this information.  This is not “cast in concrete” in the same way as Y DNA and mitochondrial haplogroups and STR markers.  Those are irrefutable – while biogeographical ethnicity remains a bit ethereal.

In summary, I would simply say that this tool can provide great hints and tips, especially the matching, which is unique, but it can’t disprove anything.  The absence of minority admixture, which is what so many people are hunting for, may be the result of the various data bases and the infancy of the science itself, and not the absence of admixture.

My recommendation would be to utilize all three biogeographic admixture products as well as the free tools in the Admixture category at GedMatch.  Look for consistency in results between the tools.  I discussed this methodology in “The Autosomal Me” series.

What Next?

I asked Dr. David Mittelman, Chief Scientific Officer, at Family Tree DNA about the reference populations.  He indicated that he agreed that some of their reference populations are small and they are actively working to increase them.  He also stated that it is important to note that Family Tree DNA prioritized accuracy over false positives so they definitely took a conservative approach.

34 thoughts on “Family Tree DNA Releases myOrigins

  1. Thanks for explain how the origins are calculated and how few it is that are in the reference population for statistics, The selection is also not random it is the people who are willing to pay 100 dollar or more to be a member of the reference population.

  2. Anyone with a long line of ancestors from Scotland and Yorkshire and even Ireland are going to (probably) show European Northland ancestry.

    Circumpolar my be an indicator of Native American ancestry? I mean because you are fairly certain of your mother’s Native American dodumentation…..

    Steve in Oro Valley

    • If you read the definition of Circumpolar in the white paper, you’ll note that it states it covers a broad region, even into the New World. My mother does have documented Native ancestry through the Acadian lines in Nova Scotia. This is documented by DNA haplogroups as well as in some records.

  3. My results are – shall we say unexpected – considering my fairly robust family tree. I have European Northlands 25% (Swedish grandfather – makes sense); North Circumpolar 4% (I seem to have a fair number for triangulated Finnish DNA cousins with no MRCA so I suspect there is a Finnish grandmother somewhere); then European Coastal Plain 68% and European Coastal Islands 3%??? Out of 16 great grandparents, one is German/Swiss, but 6 go back to Colonial New England. A few Dutch and French in the mix but mostly English, or so I thought. Of the rest, most are Irish, Scottish and recent English.

    Ancestry said I had mostly Viking ancestors. Does this mean they were Saxons and Normans instead?

    All ethnic predictions to be taken with a BIG grain of salt.

  4. Hello,
    Nice article, thank you.
    About France, you have to know that to test its own dna is…. not legal !
    That is why so few French people tried it.
    Few of them have heard about dna testing (companies don’t communicate because testing is not legal).
    We are a few to have overpassed.
    Greg

  5. Roberta – Thanks for this great post. I’ve encouraged people to read your article from my own site this morning. Overall, I’m disappointed with myOrigins, but I do see some potential benefits. The two Brazilian tribes used as New World references are just not enough, so I’m holding out hope that once they broaden their reference populations database, that the percentage figures will more accurately reflect the diverse ancestry of those of us with entirely colonial American roots.

  6. I see no benefit to myOrigins at all. Family Tree DNA has totally missed the boat with my family. I have a well-documented tree that has largely been confirmed thru DNA testing. My ancestry is mostly from the British Isles, but myOrigins has me with 0%, even though it shows my father with British Isles ancestry. But it shows my father with no ancestry in European Coastal Plain even though he has proven ancestry in France and Germany (he has more than 25% German ancestry). It shows both of us with European Northlands even though we have no Scandinavian ancestry. Also, on the Y-DNA side, our ancestry went from Spain to England, but myOrigins shows me with no English or Spanish ancestry. I have found the Gedmatch utilities to be by far the most accurate for admixture, with Ancestry.com being second best (I cannot test with 23andMe, so I cannot comment on that).

    • As I’ve tested with all three major companies, and each so different one from another, that IMHO one must receive them all with a grain of salt. Even with GEDmatch, there are several different admixture methodologies, depending on the creator, such Dienekes’ Dodecad, which has several itself based on the number of reference populations. At this stage of the science, I think everyone’s still reading tea leaves. The future will depend on a greatly enlarged database of reliable and genealogically relevant population reference samples.

  7. As usual Roberta you provide a salient summary of the latest genetic genealogy developments. One aspect of myOrigins confuses me. You say “A great new feature is the mapping of the maternal and paternal ethnicity of your Family Finder matches…” but I’m unclear what relevance a particular tester’s personal direct maternal or paternal lineage might have for their autosomal matches with whom they share no immediate family ties.

    • You are right in that the matches may not be relevant to you. But they also may be. If you see someone who has a balloon in the area where your ancestors were from, that bears checking out. It’s just another possible clue.

  8. My Origins gives me :
    78% European and 22% Anatolia

    23andme gives me 99% European and 1% Anatolia……but 23nadme state it only goes back 500 years from today, so my question is ……how far back does myOrigins cover?

    kind regards
    victor

    • I checked with Dr. David Mittelman on this question and he provided the following information:
      Anatolia & Caucasus – On the order of 5-10 K
      Asian Northeast – On the other of 5-10 K
      Bering Expansion – On the order of 10-15 K
      East Africa Pastoralist – On the order of 3-5 K
      East Asian Coastal Islands – On the order of 5-10K
      Eastern Afroasiatic – On the order of 5-10 K
      Eurasian Heartland – On the order of 5-10 K
      European Coastal Islands – On the order of 2-4 K
      European Coastal Plain – On the order of 1-3 K
      European Northlands – On the order of 3-5 K
      Indian Tectonic – On the order of 3-5 K
      Jewish Diaspora – On the order of 1-2 K
      Kalahari Basin – On the order of 50 K
      Niger-Congo Genesis – On the order to 2-4 K
      North African Coastlands – On the order of 5-10 K
      North Circumpolar – On the order of 10 K
      North Mediterranean – On the order of 5-10 K
      Trans-Ural Peneplain – On the order of 2-4 K

      • Fantastic, Roberta
        Much Appreciated

        Now to work out my
        47% European Coastal
        22% North Mediterranean
        9% Trans-Ural

        and
        22% Anatolia

        regards
        victor

      • This is very helpful, Roberta. Thank you very much for posting it. FTDNA should put this info on the explanation of each cluster on each person’s map page.

        Question – Since we all know specific DNA is in effect “diluted” the further one goes back in time… does this imply my 4% North Circumpolar would carry a lot more “weight” than say if I had 4% of European Coastal Plain?

  9. Thanks Roberta, that’s what I thought it meant. I have no words to express what I’m thinking right now. Count this as a blessing…

  10. The “Matches Map” doesn’t work on mine and I’m assuming it’s something you only get with an upgrade to testing. I transferred over Geno 2.0 results and then transferred over 23andMe when they were having a Family Finder sale a while back.

    • The matches map is only if you have tested your actual marker values – meaning the 12, 25, 37, 67 or 111 markers. It also works for mtdna if you have taken any of those tests.

  11. Hello, i’d like to inquire, which reference group is eastern afroasiatic based on? It seems quite strange that Iranians would score quite high on this. This is a question i’ve had for a long time.

  12. Roberta. The New World reference populations are Karitiana and Surui. They left out Colombians, Maya, and Pima which were are all included in Stanford’s Human Genome Diversity Project by Li et al and in the old Population Finder. You can see the breakup of Colombians, Maya, Pima, Karitiana and Surui in the Dodecad K13 by going to http://dodecad.blogspot.com/2012/10/globe13-calculator.html and clicking on spreadsheet and in Eurogenes K13 at http://goo.gl/1NqKee

    What shows up for Amerindian in Dodecad globe13 and Eurogenes K13 is similar to what used to show up in Pop Finder and it is also similar to what shows up in AncestryDNA. MyOrigins is a different story.

    MyOrigins breaks up the Native American ancestry of Latin Americans into New World, East Asian, and North Circumpolar. The largest amount goes to New World, a much smaller amount goes to East Asian, and an even smaller amount goes to North Circumpolar.

    If you have a detectable amount of Native American ancestry you should see it in Eurogenes K13 calculator at Gedmatch, Dodecad globe13 calculator http://www.y-str.org/2013/07/diy-dodecad-21-wrapper.html, or AncestryDNA.

    I assume that anybody that only has 3% Native American detected elsewhere won’t see it in myOrigins.

  13. Roberta, the way you attempted to calculate the inferred amount of ancestry of your father was done wrong. You can’t subtract your total from your mother’s total to infer your father’s total. Actually there is no accurate way to calculate the amount using what myOrigins has given you. There is a way to infer amounts based on what can be found with Dodecad globe13 or Eurogenes K13.

    I’ll explain using amounts from a family that is willing to share their results. In myOrigins the father is 65% European, 21% New World, 8% East Asian, and 6% Middle Eastern. The mother is 73% European and 26% Middle Eastern. The child is 79% European, 7% Middle Eastern, 10% New World, and 3% East Asian. See how you can’t subtract the amount of the child from the amount of the mother to get the father’s amount?

    That same family in Dodecad globe13 has the following. The father is 24% North European, 30% Mediterranean, 30% Amerindian, 4% Southwest Asian, 2% West Asian, and 10% a mix of other components. The mother is 39% North European, 34% Mediterranean, 16% West Asian, 8% Southwest Asian, and 3% other. The child is 31% North European, 34% Mediterranean, 14% Amerindian, 8% West Asian, 6% Southwest Asian and 7% other. If you notice the child is close to one half the total of the two parents. Example 24% North European + 39% North European=63%. 63% North European/2=31% which is what the child inherited. If only one parent has a component such as Amerindian then the child inherits almost exactly one half the total. Example 30% Amerindian divided by 2 is 15%. The child inherited 14%.

    Now let’s apply that to you. If you run your mother’s raw file and your raw file through Dodecad globe13 and then divide all of your mother’s results by two that is close to how much of her Autosomal DNA that you have. Now subtract the halved amount of your mother’s components from your amounts and that is how much you have inherited from your father. Now multiply each of those amounts by two which will give you close to what your father had. It’s a simple process really. Unfortunately it won’t give you the places of ancestry of your father but it will show if he had a high level of one of the components or if he is close to the European populations in the Dodecad spreadsheet.

    • I said in the verbiage that I realized that I had not obtained all of my mother’s heritage, and that my father’s would be the inferred amount or more. I also know that I didn’t receive exactly half of my mother’s – some times less and sometimes more. I worked through this extensively in “The Autosomal Me” series that I did. I have utilized the various GedMatch tools and they are great.

      • The point was that you can’t infer your father’s amount based on the results of you and your mother from myOrigins. Stating that “I realized that I had not obtained all of my mother’s heritage, and that my father’s would be the inferred amount or more” doesn’t help to explain the huge differences between the real world example that I provided and what I would get using your math.

        Please provide a mathematical equation to explain the amounts using the real world example I gave you of two parents and a child? You won’t be able to because no mathematical calculation can be used to explain the results of the child based on the parents. Likewise, the results of one of the parents based on the child and the other parent can’t be explained by using the math you are using or by any other mathematical equation.

        What mathematical equation can explain the real world example of the father being 65% European, the mother 73% and the child 79% European? Using your math I would have said that the father is 6% European. There is a huge difference between your calculated 6% and the real world 65%. The difference is so large it is meaningless to a person to know that the unknown amount of a parent could be anywhere from 6% to 65%.

        Another thing is your mother’s North Mediterranean that you didn’t inherit. You should be able to calculate that into your results and you should have close to half your mother’s results. Your math doesn’t work for that either. The math simply does not work for the results of myOrigins. It is futile to try and infer amounts based on a child and one parent from the results of myOrigins and every reader should understand this.

        I found your blog post on the gedmatch tools but it was before the newer Eurogenes K13 was released and you didn’t include your mother’s results in the blog post when you included your Eurogenes K9 results. You used a much older Dodecad which was V3. The latest is globe13. http://dna-explained.com/2013/03/03/the-autosomal-me-rooting-around-in-the-weeds-using-third-party-tools/.

        Will you do us the favor of using Eurogenes K9 or K13 with your file and your mother’s file and post your results? You will see that my math makes more sense. The numbers aren’t as close with Eurogenes as they are with Dodecad globe13 but they are a whole lot closer than they are with myOrigins.

      • Hi Armando. The point of this article was not to determine my father’s origins, but to do a quick calculation to see if the myOrigins results made sense by utilizing both my and my mother’s results. As you have pointed out, in some cases, they don’t. I don’t intend to redo “The Autosomal Me” series with updated software. That was intended to be a roadmap for people to be able to follow for their own purposes. I am very glad to see the new software both at the vendors and at GedMatch, and I may well do an article shortly comparing the three primary vendors and the GedMatch tools. As you know, this isn’t the easiest thing in the world because of the vast differences in nomenclature and trying to figure out what is an apple and an apple instead of a pear.

  14. Pingback: Ethnicity Percentages – Second Generation Report Card | 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