Skip to content

When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.

PLOS BLOGS DNA Science

Why I Didn’t Write About Elizabeth Warren’s DNA

When Elizabeth Warren released her DNA results last week, indicating a Native American ancestor a few generations back, my response was, and still is, so what?

Besides the facts that the revelation wasn’t surprising, its interpretation insulting to Native Americans, and in many accounts oversimplified or flat out incorrect, to challenge a presidential taunt (“Pocahontas”) is a waste of ATP. One of my own articles was even bumped last Tuesday, replaced with a piece claiming that “almost every American has a sliver of Native American ancestry.”

No, it’s not that simple.

But the main reason I didn’t drop everything to pontificate about Senator Warren’s relative along with the other quoted geneticists of the day is because it hit too close to home. I’m immersed in a DNA ancestry journey myself, an unexpected one that began recently when Ancestry.com discovered a close relative I didn’t know I had. And that person is just about my age – not someone from 7 generations ago.

I’m still collecting information, trying to imagine a narrative that makes sense. The memories and possibilities are coming to me slowly, as I read about other older adults whose histories may have to be tweaked in the wake of findings from direct-to-consumer DNA tests. My story will be too long for a blog post so I have to figure out where to publish it, if I publish it.

Right now, my mind is a scorecard, with reasons why the finding can’t be true on the left, yet the actual data, on the right, too statistically convincing given what I know about DNA genomes to dismiss.

As I search for answers, and inspired by the overreaction to Senator Warren’s teensy-by-comparison DNA finding, I thought it might be helpful to list some of the estimates, approximations, assumptions, and inferences that DNA ancestry testing entails. The ads are so sappy that I fear people may assume that spitting in the tube and popping it into the mail generates a genome sequence.

No, it’s not that complex.

Instead, deeming two people relatives with DNA testing is like identifying a book from a few distinctive phrases or sentences. The book is the generic genome, the signature phrases what testing actually reveals two people share.

Google “What in the Sam Hill he’s doing?” and it might not indicate much, but add “That boy’s yo’ comp’ny and if he wants to eat up the tablecloth you let him, you hear?” and a scene from To Kill a Mockingbird emerges. Scout is shocked at her friend pouring maple syrup over his dinner, when cook Calpurnia pulls her aside and explains. An incomplete scene, an incomplete genome, holds less meaning than the full story.

Still, my newfound relative and I share an uncanny number of specific DNA stretches. But most of our genomes weren’t even analyzed. Here are some of the shortcuts:

  • Ancestry testing to match relatives considers less than .02% of the genome, in the form of 700,000 or so SNPs (single nucleotide polymorphisms) and not always all four DNA base possibilities at each site. That’s 700,000 out of 3.2 billion bits of DNA information.
  • The algorithm that spits out cousins, aunts, and grandparents is trained on pairs of people who are related in known ways, like aunt/uncle to niece/nephew or first cousins. It’s searching for a machine-learned pattern.
  • Also considered are the frequencies of short DNA sequences in populations around the world. But not all populations have been analyzed, and companies’ databases may disagree or be proprietary.
  • The pie charts of geographical origins assume that grandparents and relatives farther back didn’t move around much. Like assuming most of us have a smidgeon of Native American DNA, that simply wasn’t always true.

The larger lesson: DNA testing, of any kind and especially direct-to-consumer, is serious. Not entertainment. Think before you spit, because once you know something, you can’t unknow it. I wish I’d never sent samples to companies offering to test me for free if I wrote about them, years ago. After I’d learned I was 100% Ashkenazi, which I already knew, I’d forgotten all about it.

When I wrote “DNA Testing Kits as Holiday Gifts Can Bring Surprises” here last December, I had no idea it would happen to me.

The bottom line is that until a few weeks ago, I thought I knew everything there was to know about DNA testing. But I had no idea of the sucker punch that unexpected findings can deliver – even when they’re good.

 

 

 

 

 

 

 

 

 

Discussion
  1. Hi, I am a bit puzzled by the line: “Ancestry testing to match relatives considers less than .02%”
    But if this 0.02% is evenly distributed, why wouldn’t it be meaningful? Why is this considered a shortcut?

  2. Why does revealing the unknown have to be bad? DNA testing showed an un-known cousin for both my mother and father. Awesome! The family has expanded to whatever degree that other person chooses. Our parents, grandparents, etc., and their siblings were/are not perfect, and neither are we. Accept that our family members are not infallible and embrace the new information.

  3. I appreciate what you say. I first tested ftDNA in 2005, helped sponsor a kit the year before that and can’t tell you where all that has taken me. Helping many people identify the very relatives you mention, prove believed lines and blog and support several dozen private family groups for genealogy. I also got the sucker punch myself in July and there is no explaining and I am still with dropped jaw. But please let me add a note about Elizabeth Warren and folks in her position. There is likely good reason that countless families have “the story”. We even have those small amounts on my father’s side with is almost all that US southern which has everything on earth including 0.001% martian – the melting pot people. And for my mother’s well that is another story, yes she has those traces, hers is Quebec and it is mixed with racism and painful herstory. But Senator Warren’s mother came by the story about the Smith Native American heritage honestly. I believe if we have any complaint we need to have the online records and tree companies like ancestryDOTcom to not just put leaves on these folks but put red flags that this information may or may not be true. Of course this is what is said by the companies already but how many of anyone read all the fine print. The story of this Native ancestor has been circulating a long long time. From some of the references on the sites, I can’t see why the story would have been rejected and so there was every reason to believe it. Maybe someone should devote time to proving or disproving. But the genetic genealogy writers I have trusted for ages now place Ms Warren’s percentage at about the amount it should be for that distance of ancestor. So it certainly can’t be disproved. And there were 1,000s I bet hoping she would test. I don’t think Ms Warren did anything wrong and I don’t think it was disrespectful to the Native community. If we think changes are needed there I think the source would be going to ancestryDOTcom, rootsweb national archives archives in Quebec and remove all the records of people who would only be a tiny percentage of anything? No a bit harsh

  4. It is very meanningful, I didn’t mean to imply that it isn’t. It’s just not complete, and I fear some people might assume that entire genomes are sequenced. 700000 SNPs across the genome is a lot of information. But if, for example, my newly-found relative and I had different alleles at the CFTR locus, for example, that wouldn’t show up, because it isn’t part of the test.

  5. It is just my opinion based on what I am feeling now. The new knowledge of a close relative is fantastic! But I’m rethinking some weird things from my childhood that now make sense, and it makes me sad to know some of the problems my parents (who have passed) had. They kept it from me.

  6. Thanks Cherie Lynn. Re Elizabeth Warren, mostly I objected to her trying to answer Trump, because not he’s off calling her names all over again. It is almost disappointing to be Ashkenazi because we are all so alike.

Leave a Reply

Your email address will not be published. Required fields are marked *


Add your ORCID here. (e.g. 0000-0002-7299-680X)

Back to top