When genomic ancestry tests lead to confusion

3 minute read

An article in Gizmodo by Kristen Brown asks an uncomfortable question about today’s proliferating genomic ancestry industry: “How DNA Testing Botched My Family’s Heritage, and Probably Yours, Too”.

Her family thought their ancestry was mostly Syrian, and they were surprised to get a very different assortment of results from different genome tests for ancestry. That’s not unusual, it happens a lot.

Brown accentuates the explanation that the tests seem to work better for people with predominantly European ancestry because the comparative samples are much larger and pose a limited set of problems compared to other regions.

In my opinion, that’s really not true; it’s just that the issues appear on a different scale. People who derive their ancestry from different parts of Europe are often given results that don’t correspond to their genealogical history.

The article does a good job of discussing the way that people’s expectations of these tests doesn’t match what the scientists are trying to provide.

A big problem is that many of us have a basic misunderstanding of what exactly we’re reading when Ancestry or 23andMe or National Geographic sends us colorful infographics about how British or Irish or Scandinavian we are. It’s not that the science is bad. It’s that it’s inherently imperfect, an estimation based on how much our DNA matches up with people in other places around the world, in a world where people have been mixing and matching and getting it on since the beginning of human history.

But right now the science really is bad.

To me, the great promise of personal genomics is that people can advance the science by engaging with their place in this history. Naturally people are most interested in their own places in the human genealogical web. Anyone can be forgiven if they see this genealogical web as a tree of populations with names like French, German, Italian, or Hausa, Bengal, or Syrian.

So a person’s place in the history seems like it should be some combination of these names, like a pedigree.

It’s very easy for genetics to deliver a combination of names. Trouble is, a single combination of genes may map pretty closely to several different combinations of population names.

Geneticists study samples of human genomes from many different parts of the world as a basis for these ancestry assessments. Those samples of human genomes are simply too small in most regions of the world and most ethnic groups. As much as we know, we are still learning about how today’s groups have originated, and today’s nation-states have come together from ethnic groups in many different ways.

Samples are of adequate size in much of Europe, but still, humans across hundreds of years have not behaved in ways that make it easy to reliably apportion ancestry into geographic bins.

In other words, our knowledge of recent human genetic history is a work in progress. Any one individual’s place in recent history is contingent on the whole story. We know some parts better than others, but even the well-traveled parts of human genetic history have hidden chapters.

Populations did sort of behave like a tree during much of human prehistory–but a tree with millions of interconnections between branches. Anthropologists and geneticists are trying to reconstruct that tree with greater accuracy. They are also using ancient DNA to rediscover branches that are missing from conventional histories.

Braided stream
A braided stream. Photo Credit: Sam Beebe, Ecotrust via Compfight cc

Even though individuals belonged to populations, the genealogies of individuals do not reduce well to a tree of populations. Populations are loosely defined by swarms of individuals with similar genealogies for a few hundred years. Any one individual’s genealogy may diverge quickly from the mainstream of the swarm.

I’d like to promote a broader view of genealogy, one that celebrates the differences between individuals and the populations that some of their ancestors identified with. That has more genetic reality than the alternative, in which people are defined into hardened populations.

See also: “The surprising connectedness of human genealogies over centuries”