AncestryDNA | Breaking Down the Science Behind Ethnicity Results | Ancestry

Join Ross Curtis, PhD from the AncestryDNA science team as he talks us through the science behind the ethnicity results. Learn how we determine the ethnicity estimate and why it can be different than a sibling.

Start Your Journey Today:

About Ancestry:
Bringing together science and self-discovery, Ancestry helps everyone, everywhere discover the story of what led to them. Our sophisticated engineering and technology harnesses family history and consumer genomics, combining billions of rich historical records and millions of family trees to over 10 million and counting to provide people with deeply meaningful insights about who they are and where they come from.
We’ve pioneered and defined this category, developing new innovations and technologies that have reinvented how people make family history discoveries. And these discoveries can give everyone a greater sense of identity, relatedness, and their place in the world.

Connect with Ancestry:

AncestryDNA | Breaking Down the Science Behind Ethnicity Results | Ancestry

[puncheex]

And now for some of the history that backs up the science:

In ancient times humans were hunter/gatherers, mainly. That ended when agriculture as a way of life was discovered, in the 5, 000 years before today. Farms tend to tie people to the land; they are possessive and protective of it, and they settle. Even after domestication of animals allowed for movement on horseback, it was rare for an individual to travel more than 20 miles away from their place of birth, and that applies even more so to the seeking out of mates. Natural geographic barriers, like oceans, rivers and mountain ranges tend to define what this 20-mile radius meant in practical terms, at various nesting levels of size from towns to pre-nations.

In this sort of environment the indigenees tend to become somewhat inbred. Not in any pejorative way most here well understand, mainly, but in the concentrating of traits useful to people in the area. The people around the Baltic Sea develop what might be called super-white traits - light hair and eyes that stay through a lifetime rather than fading in adolescence, for example. These traits are useful in an area that wouldn't be able to support much farming without the warm Gulf Stream currents that feed the area, but with them support a large farming population in an area of weak sunlight and the vitamin D and folates deficiencies which result. This is the same sort of situation which would lead eventually to the speciation of H. sapiens, but there was enough wanderlust even in those times to keep the groups from separating genetically to that extent.

Until around 1750. About that time better transportation systems, the opening of new areas to be settled, better education generally available about those areas, and the weakening of feudal systems and other political barriers allowed for a more general mixing and movement of individuals, breaking up the old groups.

The science which explained all this in genetic terms only became available in the late 1900s. Some smart cookie realized that if they could develop a way to read a person's genome (or even the small percent of it containing the significant data, called "markers"), they could develop a database of the markers which were boosted or shunned within the early groups, and match the two to be able to determine which groups a person's ancestry came from. Ten years ago this "autosomal" testing (meaning all ancestry, not just the patri- and matri-lineal lines analysis also known as genetic testing) could detect only three groups: European, African and native American. The database grows larger with every test, though, and more sensitivity or more accuracy are the result.

Caveats - All this testing depends on the companies' proprietary databases of previous tests and of the markers found to be more powerful group indicators - there are 3 billion possible markers, so lots of room for finding good markers exist). The results will vary over time and between companies, and include rather wide error ranges. The "groupings" were mainly on geographical areas, not the nations we recognize today - for instance, one grouping is the Iberian peninsula, which ignores the differences between Spain proper, Portugal, Catalan, Castilian and so on. Any grouping formed late would have less chance to develop the genetic markers necessary to this testing, so there is no "America" group, other than American Indians; Mexicans and Brazilians are not noted as a group, though Iberian and American Indians are. Note that "race" is not a concept that even fits with the ability to do autosomal testing; race is not well enough defined in a scientific sense to be an aspect of reality, let alone play a role.

These historic underpinnings will not occur again until planetary (at least) space flight becomes common, and only then if something allows for the long time isolation on planets as in our terrestrial case. So autosomal testing is only available for a set of groupings which are fading farther into the past with every new day. There is far too much travel and mixing, not enough evolving, today to ever be able to develop autosomal testing 300 years from now based on where people were in 2000.

Wow. Did I get carried away!

[alpacino]

Oman shows up on my central asia region, but isn't it supposed to be a part of middle east ? Please help me in this regard

[NCandHR]

@Ancestry Since this video is 5 years old, are you still using your inital reference panel or are you now including other folks (like those who submit DNA) who have lived in the same european region for generations? Is the test more robust now than it was 5 years ago.

[thebeautifulgame]

Hi...As your reference panel seems key to all this, are you actively expanding the "reference panel" in order to make it more precise & reliable (eliminate distortions from smaller sample sizes).  If yes, how much has it grown in percentage terms?  Thanks!

[yminoh]

Will you ever be able to break down 'Europe East' or 'Europe West' instead of it lumped with several areas?

[peterbaxter]

It becomes more and more obvious that each company uses a different set of algorithms to calculate ancestral DNA - thus it is little wonder that results differ from company to company. I have now tried FIVE, none of which bears a great deal of resemblance to any of the others.

[stingrayofcincinnati]

Yminoh, I am totally with you on that.  Most of my family lived in Eastern Europe.  Need to format the old hard drives and send the computers to Eastern Europe.  They need help...lol

[dreamrecaller]

I got my DNA Results back 100% European! Very interesting

[exocet]

Has anybody ever thought about the idea to come along the family tree from the other side? I mean by taking samples from sites of archaeological cultures such as the Yamna Culture and comparing them to modern day samples?? I think that would make much more sense than giving the people percentages of Nationalities who have nothing to do with DNA but are more artificial creations, just think about being told you are 22% English eg. if you go further back you'll end up back in Germany where they originated, also you are just comparing them to people whos families ALLEGEDLY have lived there for a longer period, so that all absolutely doesnt make any sense. You pretty much have to do it the other way around in order to be somehow scientific.

[ArielPontes]

How is the percentage of Amerindian ancestry calculated? How was the reference point established? I can go about 5 generations up in my family tree and everybody was born in Brazil. However, I believe I have quite little Amerindian ancestry (as most average Brazilians). So if a person like me was used as a basis to establish a reference point of "average native Brazilian" just because my ancestors have been in the region for many generations, the results would be pretty unreliable. This method makes sense for Europe or "the old world" in general. But for the Americas and Australia, how are the "native reference point" established?

[joshou]

uhm how expensive is this in Philippine pesos just asking I'm interested in learning my lineage.

[MeLancer]

Thank you for the explanation. I've used Ancestry DNA for a few year now and appreciate the revised estimates you publish from time to time. I assume other companies use slightly different algorithms than Ancestry. I recently uploaded my Ancestry RAW DNA to FamilyTree DNA. In Ancestry, all my ethnicity is from the British Isles and NW Europe. FamilyTree DNA shows a southern Europe component from Italy to Greece and the Balkans.

[Angbwillinspireu]

Thank you! This video was super helpful.

[marisabasso]

Hi, so i got my test, mailing it in. will i get a hard copy if the results? how accurate us the test.how do you go back thoysands if years if you dont ohysically have dna from 100s of years ago.

[TaleofTwoSiblings]

just curious do all the dna tests go to dubin ireleand or is it just the ones that are not from the USA ?

[roxychic]

Just wondering the test I did with ancestry DNA will pick up on native Australian DNA?
I hope you guys looking into native Australian DNA groups if you have not already

[MichelleDespres]

What do the small black numbers next to regions under "see all regions" mean?

[kairishahad]

How can you possibly say that people with known ancestry (grandparents, great grandparents) are indicative of the DNA of that population, when in every FAQ page and video, you always recognize one of the downfalls of atDNA is that you don't inherit DNA from every ancestor, and that it's a randomized process? Let's say just two of their great grandparents' parents were from Africa, but the child was born in Norway. That's going to result in 1/4 of the available DNA for that person to inherit being from Africa, and for all we know they very well could have inherited all 25% of it.
Using such small sample sizes to represent millions of people is simply illogical, especially when you use 131 individuals to account for the DNA of over 500+ culturally, linguistically, and ethnically diverse nations.

[zoniahill]

How long a period have you all been researching people's DNA to attain this information of their ethnic origins?

[marcelosurville]

Does anyone know if Ancestry.com tells you to what Haplogroup you belong to? please