film / tv / substack / social media / lists / web / celeb / pajiba love / misc / about / cbr
film / tv / substack / web / celeb


Don’t Fall for the Impossible Promise of DNA Ancestry Kits

By Ciara Wardlow | Miscellaneous | June 23, 2020 |

By Ciara Wardlow | Miscellaneous | June 23, 2020 |


As a Black woman descended from enslaved people, I often see ads for DNA ancestry kits that promise they will connect me to my roots, and I feel sorely tempted. Where in Africa were my ancestors stolen from? What languages did they speak, what beliefs did they hold? Could I somehow recover even the tiniest little piece of those identities stolen so long ago?

It is not a yearning exclusive to the Black community by any means, considering the various populations across the globe that have been forced from their ancestral lands and/or had their heritage and culture forcibly suppressed throughout the ages, but it is particularly common to the Black community. While I don’t have hard statistics, from personal experience and anecdotal evidence I do feel that at-home DNA kits have been targeting Black consumers on the ancestry front more and more, realizing the particular allure their claims have for the many descendants of enslaved ancestors stolen from their homelands and purposefully stripped of their identities—their beliefs, their families, even their names. I’ve started hearing more friends and family talking about using at-home DNA kits; I see more and more Black actors in DNA kit advertisements; at a Juneteenth celebration I attended one speaker openly encouraged the practice, enthusiastically sharing her own results.

I understand the yearning. I feel it myself and I know both sides of the coin—I know exactly the sense of place and heritage I’m missing on my father’s side because I have it on my mother’s. While my father’s ancestry is full of questions and a surname that once belonged to a white man who owned my forebears, my mother is Irish, born and raised in a small town on the same land her family has farmed for generations. I spent most of my childhood summers there, exploring the dilapidated farmhouse that started off as a one-room stone cottage predating the Irish Potato Famine, hearing about how (for reasons I have never uncovered) my ancestors fared better when the blight came than several other nearby tenant farmers, who, so the story goes, would rummage through our garbage heap in search of turnip peels.

There is a sense of belonging and legacy in knowing a place like that, in being able to go to exactly where your ancestors lived and died, to sit in the shade of a tree you know a forebear planted. It is a privilege so widespread in some communities they have little notion of its specialness, or how many people have had that privilege stolen from them, sometimes several times over.

The prospect of being able to reclaim some of that privilege through genetic ancestry testing is hugely alluring. Unfortunately, it is also too good to be true, a promise that cannot really be kept, a stack of half-truths and statistical guesswork presented under a false veneer of certainty.

Consumer genetics companies sell us all a narrative—a narrative of ourselves, of science certified self-discovery. It’s a very tempting narrative, and there’s now a multi-billion dollar industry that indicates they have gotten very good at selling it.

Most of these companies really sell three different things—ancestry, health information, and genealogy (i.e. connecting family trees)—and sell them as if all three are equally accurate and scientifically backed. This is absolutely false. If connecting relatives and flagging genetic alleles of potential medical concern is like going outside and seeing what the weather is like, attempting to trace one’s ethnicity and ancestral heritage via genomics is a lot more like looking at the tail end of a 30-day forecast. There’s a scientific basis to the projection at some level, but also significant room for error and, in the case of ancestry testing, lots of suppositions at play that can never be fully verified.

There’s been some reportage in the past few years regarding how DNA kit consumers have actually seen their ancestry results shift enormously as companies revamp their predictive models, but not nearly enough discussion of what that means, and exactly what it says about the half-truths and misconceptions on which this hugely profitable sector of the genomics industry is based.

Sure, companies lie through their teeth in advertising all the time, but there’s something particularly disgraceful and infuriating about building an entire business around the impossible promise of revealing the “truth,” so let’s break it down.

No matter how tempting the “find your heritage” message may be, or how adept DNA testing kit companies have gotten at pushing it, the thing that always stops me from succumbing to the siren song is a memory of a single PowerPoint slide from a Genetics course I took in college.

There’s a concept known as gene flow that refers to the transfer of genetic variation between different populations, and on that particular slide was a maxim so popular it’s often abbreviated: One migrant per generation is enough to conserve gene flow between populations, or OMPG. In other words, if there is any interbreeding at all between two groups, they are not going to diverge in any genetically distinguishable way. To apply this to the big picture, all human beings are 99.9% identical, and the variation in that remaining 0.1% really doesn’t break down in a way that will tell you much about an individual’s ancestral origins since our species, generally speaking, has long had a propensity for moving around and procreating with other human populations encountered along the way.

At this point, you might be wondering what ancestry DNA tests are actually testing, then, and how exactly they generate all those impressively specific percentages and spiffy graphs they give you in your results. The answer to that involves a whole lot of statistical calculations and banking on frequencies.

Here’s a quick, hugely simplified run-down: there are regions within the human genome known as single nucleotide polymorphisms, or SNPs (“snips”). These are basically hotspots for genetic variation, places where it’s particularly common for random mutation to lead to one nucleotide being swapped out for another—a C to a T, for example—in a way that usually has no impact on an individual’s health or development but is from there passed down through the generations (unless, of course, another mutation occurs somewhere down the line). SNPs are, in other words, silent mutations, but they are the lifeblood of ancestry testing because having a lot of SNPs in common suggests shared ancestry. The more SNPs in common, the more recently individuals likely shared a common ancestor. As for what “a lot” means in this context, of the 3.2 billion nucleotide base pairs in a human genome, some DNA ancestry tests analyze half a million or more SNPs.

As described earlier, there aren’t specific Somalian or Nigerian or Irish or Italian genes to look for in the DNA itself. What these companies do is take “known” reference populations to identify SNPs and patterns of SNPs that are particularly prevalent within particular ethnic groups and then analyze a customer’s DNA for the presence of these.

So, where do these “known” reference genomes come from and how can we all be so sure these people accurately know their ancestry dating back hundreds of years when the rest of us are being marketed kits to figure that out? The answer is we don’t know, and these companies are incredibly tight-lipped about the whole thing.

I tried clicking around the 23andMe website for a while and eventually came to a very long, graph-filled page about the company’s “state-of-the-art geographic ancestry analysis” which quickly boasts about the “14,000 people with known ancestry” providing the basis for their analysis. The page then spends several thousand words giving an introductory genomics lesson that manages to not address how their analysis actually works at all, or why anyone should actually put any stock in the results it spits out. When the foundational reference data-set is brought up again, it is described as including “genotypes from 14,437 people who were chosen to reflect populations that existed before transcontinental travel and migration were common (at least 500 years ago),” which really, if anything, just begs more questions instead of providing any meaningful answers.

On the one hand, if you really believe in the unparalleled integrity of large corporations, you could make the case that such obfuscation is necessary to protect valuable trade secrets and there is nothing more to it than that. On the other, there is room to suspect between the handwaving and the support pages with fine print disclaimers like “predicting inheritance is not an exact science” that the secrecy might also have something to do with not wanting potential customers to cotton on to the fact that they’re selling a service that is one part science and two parts snake oil. Regardless, there’s a distinct irony to having a whole industry built on the assumed accuracy of a “known ancestry” reference pool that most frequently promotes its product with ads about people thinking they know their heritage and discovering they were wrong.

Genetic testing isn’t fundamentally evil or bad. There is a long list of privacy and ethical concerns surrounding the practice that have worryingly not yet been addressed in any meaningful way, but there are a number of legitimate applications for genetic testing that have the potential to do a lot of good if regulated appropriately. I, myself, have taken a particular kind of genetic test before and found it helpful. I deal with depression, and when trying to find a medication that was a good fit, my psychiatrist at the time suggested I do pharmacogenomics testing, or drug-gene testing, which specifically looks at your genes for variations known to be correlated with limited efficacy or experiencing side effects from a range of medications. My results indicated that the mechanism of action at work in many common antidepressants would be likely ineffective in my case, so the psychiatrist ended up prescribing a relatively new drug that worked differently. It ended up being a good fit for me. Odds are she would have suggested something else if I hadn’t done the testing, and I ultimately think the test was worth it.

Genetic testing is a powerful tool, and only growing more so as our understanding of genetics evolves and methods become more and more sophisticated. But genetic testing is not all-powerful. There are certain things it cannot really tell us, not because we need bigger data sets or are still waiting on crucial breakthroughs, but because the shoe just doesn’t fit, and it never really will.

Here’s the thing about race, ethnicity, and culture: it’s not genetic. Yes, when you look at things like skin color and hair texture, there are connections, but it’s a Venn diagram with limited overlap. That’s not to say that race, ethnicity, and culture are somehow not real or not important. They are all very real and incredibly important. They just are not ultimately genetic—although sometimes genetic-adjacent, if you will.

“We’re clear upfront that DNA is not identity. DNA is not culture,” Robin Smith, the head of 23andMe’s Ancestry Division told STAT News for an article published last year, although clearly not the company’s marketing department. Like other DNA ancestry kit companies, it rakes in millions of new customers every year with the allure of “discover who you really are,” not “we’ll send you a guess of who we think you might be based off of contested hypotheses and dodgy methodologies, then sell your genetic data to third parties for hundreds of millions of dollars” because somehow people might not be as interested in forking over $100 or so when presented with the second option.

There are a lot of stories out there—some heartwarming, others heartbreaking—about DNA kit results completely reshaping a person’s sense of self and family history. Now, if the results tell you that you’re positive for a BRCA1 mutation or you have a long-lost cousin or your uncle is actually your dad—that’s all relatively straightforward, cut-and-dry genomics, and unless a lab technician processed your tube of spit wrong somehow, the findings are almost certainly accurate. But ancestry is a different matter. If you want to do a DNA ancestry kit as a bit of a lark or out of idle curiosity, that’s one thing, but know that whatever results come back are just guesses, and can never be anything more.

Header Image Source: Science & Society Picture Library / Contributor