This is Part 4 of a multi-part series, The Autosomal Me.
Part 1 was “The Autosomal Me – Unraveling Minority Admixture” and Part 2 was “The Autosomal Me – The Ancestors Speak.” Part 1 discussed the technique we are going to use to unravel minority ancestry, and why it works. Part 2 gave an example of the power of fragmented chromosomal mapping and the raw beauty of the results. Part 3, “The Autosomal Me – Who Am I?,” discussed how to use our pedigree charts to gauge expected results and how autosomal results are grouped into population buckets. We also named this technique, Minority Admixture Mapping, or MAP for short.
In this segment, Part 4, let’s take a look at what the testing company autosomal results look like. The results are presented in timeline order, with the oldest results first and the latest, and presumably most accurate results, last.
23andMe Version 1
23andMe was the first company to offer this type of testing affordably. They initially only offered 3 population groups, and one inferred that Asian was actually Native American. Of course, that wasn’t a valid assumption for everyone, but it was the best that could be done under the circumstances. This was my ethnicity results display at 23andMe until December 2012 went their updated version was released.
DeCode Genetics initially offered autosomal tests for ancestry. Unfortunately, under the pressure of financial issues, they stepped away from the genetic genealogy marketspace and have since been sold.
Their test showed the following ethnic breakdown, picking up both my Native and African heritage:
I particularly like these results because the X chromosome is included, and seeing Native on the X chromosome, which has a unique inheritance path is a very important piece of data.
Family Tree DNA Version 1
Family Tree DNA’s first version of their Family Finder product produced results stating that I am 100% European, split between western and northern, shown below (minus the map.)
Dr. Doug McDonald
Doug McDonald, a retired physical chemistry professor, compiles contributed raw data and compares the raw data locations with both reference populations and the contributor results. This is not a commercial endeavor but a private research project which has been ongoing for years. His analysis of my raw data results from 23andMe and Family Tree DNA showed that they are primarily European. His first analysis was without Middle Eastern populations and the results showed European except for a total of about 3% East Asian, Oceana and American. However, in a second run including the Pakistan and Middle Eastern populations, the results now showed 88% European, about 1% Oceanic and American and the balance Middle Eastern and Pakistani.
A small amount of Middle Eastern heritage is not unexpected since I do have confirmed Turkish ancestors.
Dr. McDonald indicated that this was slightly more, 1-2%, than most Europeans, and that I was generally planted firmly in the middle of the “English” area in his data. His results showed no African.
Standard deviation (statistical noise) is about 1%. He can achieve these low deviation numbers by using such a large number of markers (536,904 to be exact) for his comparison. I am grateful to Dr. McDonald for his contribution, not only to me, but to this field.
The graph below shows that my primary ancestry falls in the English/French region.
The second graph maps these results on my chromosomes. The American, bright green, is found on chromosomes 1 and 2, and the X chromosome shows South Asian.
Doug indicates that the Native American is found at about the .5% level. Interestingly, on my mother’s graphs and charts (below), the Native segments are nearly identical, but my first grey South Asian segment on my X is Mideast on her chart.
It’s also interesting to note that my Native American on chromosome 2 is larger than my mother’s which may well reflect Native heritage on my father’s side. Ironically, the oral history of Native ancestry was on my father’s side, not my mothers.
Doug’s analysis has been updated several times over the years and these results are the most current. The vendors have made upgrades too. In 2012, both 23andMe and Family Tree DNA underwent upgrades to their ethnicity software and the Genographic Project version 2.0 test was released.
23andMe 2012 Updated Version
The new 23andMe software offers different confidence levels.
The standard estimate, or confidence level, shows that I have about .5% Native American. This is consistent with Dr. McDonald’s findings.
A second view is available which paints the chromosomes. A split view is also available if one of your parents has been tested at 23andMe as well. That is not an option for me.
The conservative estimate, below, shows less Native at .2%.
The speculative level below shows the Native back to .5% but adjusts the European regions significantly.
Although 23andMe does not provide participants with the start and stop locations, through alternative means, meaning a very smart friend, Rebekah Canada, who is a Java programmer, start and stop locations can be discerned.
CeCe Moore documented Rebekah’s technique for those who will be following along with their own results through this process.
In a future segment of this series, we’ll look at alternative ways to discern Native segments. Thanks to Rebekah’s technique, I can tell you that 23andMe shows my Native segments as follows:
Chromosome 1 – 165,658,091 to 175,711,116
Chromosome 2 – 86,316,174 to103,145,426
23andMe also provides a Neanderthal percentage. What fun!!!
Family Tree DNA Updated 2012 Version
My mother was deceased before chip based autosomal testing was available, but I ordered the Family Finder test for her as soon as it was available. Thankfully her DNA was stored at Family Tree DNA and was still viable.
Mother’s original results are shown above and her most recent results are shown below. Her results shifted within Europe and her margin of error doubled.
My current results from Family Tree DNA’s updated software are shown below.
National Geographic Genographic 2.0
I was very surprised to see my National Geographic results. They were very unexpected, in particular the high percentages of Mediterranean and Southwest Asian, totaling 54%.
It made more sense when I read the information. It’s true, reading is fundamental.
These results are, in essence, more anthropological in nature.
Of course, one of the fun parts of the Genographic results are the Neanderthal and Denisovan percentages.
These are somewhat different than the 23andMe results, although if you add the Neanderthal and Denisovan values together, the resultant 2.2% is very close to 23andMe’s 2.5%.
In 2012, Ancestry introduced an autosomal DNA test as well. What it provides is very limited, with limited tools, but it does provide percentages of ethnicity in addition to matches. Recently, Ancestry announced that the percentages may change over time. They have been severely beaten within the genetic genealogy community for quality issues with this product, including percentages of ethnicity that are highly erroneous. Their stated time reference is 500 years ago.
Recently this new page was added before you can see your detailed results.
Ancestry shows my heritage as only British and Scandinavian.
Ironically, Ancestry has mapped the birth locations of my ancestors in Europe on the map above, based on my family tree submitted. Interesting that Germany doesn’t show in Ancestry’s ethnicity list but many of my family lines originated in Germany and Holland, and none in Scandinavia.
Testing Provider Summary
Where do we stand now?
A summary of the various test results is shown below compared to my pedigree analysis.
Test Results Chart
I have included Dr. McDonald’s analysis here, not because he’s a testing provider in the sense of the testing companies, but because his offering was available in this timeframe, and because he worked with Family Tree to develop their Population Finder code.
You can see that the results are relatively consistent between testing companies. There is certainly no question about majority ancestry, but the minority admixture which hovers someplace near 1%, give or take 5% in either direction, is much less consistent and not always reported. If I were to have tested with only one company and taken the results as gospel, I could certainly have been left believing that I had no Native or African admixture. For many people, it’s this small amount of minority admixture that they are seeking. So in answer to the question of which testing company is “best,” the answer is, if you’re looking for trace amounts of anything, the compendium of all the testing companies (minus Ancestry) would provide the best set of results. We will be using the match information as well in the next sections, so certainly nothing has been “wasted” testing with multiple companies, again, except Ancestry. I am hopeful that Ancestry will in the future release our raw data (which they have promised to do) in a useable format, fix their misleading ethnicity results and add chromosome painting tools so that we can fully utilize our data.
In Part 5 of the series, we’ll take a look at third party tools and how they can continue to refine and add to our knowledge of our admixture.
1. Genealogy-DNA Rootsweb posting by Doug McDonald on 7-26-09 and personal correspondence.
2. 71.5% western European, 28.4% Northeastern European
3. Inferred that Asian is actually Native in an American with no history of Asian ancestry.
4. No category, inferred.
5. 78.6% Northern European, 1.8% Southern European, 18.7% Nonspecific European
6. 54.6% Northern European, .3% Southern European, 43% Nonspecific European
7. 91.7% Northern European, 3% Southern European, 3.3% Nonspecific European
8. 75.18% West Europe (French and Orcadian), 24.82 Europe (Romanian, Russian, Tuscan and Finnish). Note that my mother’s results are almost identical except the Finnish is missing from hers.
9. 43% North Europe and 36% Mediterranean
10. 80% British, 12% Scandinavian