People love to be told what tribe they belong to, I guess

I submitted a DNA sample to 23andMe, and now they send me periodic updates on what they’ve figured out lately from my genes. For instance, they’ve informed me about traits I might have.

This is strange. I know what I take away from it, but I wonder what the general populations learns from this kind of list. Here are the two important messages I learn from this kind of revelation:

  • Notice all the “likely”s and “less likely”s? That will never change, because if there’s one thing we can be confident of, it’s that every trait is the product of multiple genes interacting in complex ways. It says I’m not likely to have a cleft chin, but I know for a fact that under my beard there is a cleft chin, and that I’ve passed that on to my children and inherited it from my father. It doesn’t mean 23andMe did something wrong, it means that they’re missing information about all the factors that contribute to chin shape, and are estimating from knowledge of a few genes. They’re drawing on correlations in the large database that they have, but can’t infer mechanisms or the full range of interactions that occur during jaw development.
  • What’s really depressing, though, is how trivial all these traits are. Does anyone really care that I’m less likely to get dandruff? Wouldn’t I get better information about that by checking my shoulders than by analyzing my DNA? Part of the reason for the triviality is that they also have a “health report” that they charge extra for which summarizes more substantial predilections, which I haven’t paid for. It would have exactly the same kind of probabilistic statements. It might, for instance, say I’m genetically predisposed to heart disease, but I could probably guess that from the fact that my father died young of heart disease, and would be better off going to my doctor to have my cholesterol and blood pressure checked (as I do regularly, and no worries — I’m in good shape for a man of my age).

So what I take away from this is a lesson in uncertainty and doubt; is that the information they’re trying to share with the general public? Because that’s not what I get from their “traits tutorial”, which starts off “Our Traits Reports are a fun way to explore how your DNA makes you unique.” Where’s the fun? Where’s the uniqueness? Am I special in my resistance to bunions, or am I supposed to be entertained with the news that I probably have blue or green eyes? (Not blue, by the way, more green, kind of hazel; I usually say they’re the color of rich algal swamp mud).

My general impression is that this is “fun” in the way taking Myers-Briggs tests and horoscope readings are fun; it’s mostly bogus, phrased in a way to seem positive, and we can poke through them for affirmations of stuff we were pretty sure of, already.

I signed up mainly for the ancestry component, but even there, it’s vague.

My god, I’m a white guy! Who’d have guessed?

Meanwhile, this is just nonsense.

Yes, I’m a member of haplogroups that include European royalty, which is true of almost all the white people from Northern Europe. You might as well announce that I have pale skin, just like the kings and queens of old Europe! Whoop-te-doo! I am fun and unique.

I am not opposed to the idea of 23andMe, and think they’ve gathered a lot of potentially useful information. I just feel that the way its presented to the public is biased to reinforce false ideas of genetic determinism to induce people to participate, and that worries me.


  1. anat says

    The health part was informative in some ways, especially wrt drug response – my medical chart now has warnings about potentially higher sensitivity to one common drug and potential low responsiveness to another. I also learned about risk for certain conditions with major predisposing alleles and am working on prevention.

    But what was even more informative was getting my whole genome sequenced and getting polygenic risk scores for many common clinical measures. For instance now I know that unless I keep exercising vigorously my HDL levels are destined to be on the lower end.

  2. Kevin Karplus says

    The 23andme report is more cautious (and probably more accurate) than the reports from sequencing services. The health report mainly says “variant not detected” for 44 monogenic hereditary diseases (an easy test) and “variants not detected” for a dozen more complicated risks (with “slightly higher risk”, “variant detected, not likely increased risk”, and “typical likelihood” sprinkled in).

    With whole-genome sequence data, I could get a much more detailed report (with a lot of much harder to interpret parts) from Promethease using SNPedia. Promethease will also interpret 23andme data using SNPedia, as well as looking for conflicts if you have multiple sources of DNA variant information for the same individual. Promethease just sold all the genome data they collected to MyHeritage, and any future submissions will appear in the MyHeritage database.

    I think that a lot of people get 23andme kits mainly for the genealogical information (the list of DNA relatives). The haplogroup information is probably of little use to them (though 23andme does narrow it down more than some other services, identifying a paternal haplogroup that is only about 17000 years old.

    Somewhat interestingly, all the1225 relatives that 23andme found for me were on my Dad’s side of the family, none from my mother’s side. I don’t know whether this is because that family is larger, more likely to use 23andme, or more genetically distinctive and so easier to identify as relatives.

  3. bowdsquared says

    I don’t use it for the health data, but the cool thing for me is seeing where on the globe my genes cane from. I found the European ancestry that I was expecting, mainly England and Germany but there were some neat surprises. One was Italian genesis showing up which I have no idea how they got there, and Native American which just like the Italian I am wondering what it was doing there. My mother picked up some from Africa which was neat to see. My daughter lights up just about everywhere except Africa, Australia, and the Middle East.

  4. says

    Imagine our amazement when our two Han Chinese daughters turned up 90+% Han Chinese.Plus another 1.5 to 2% Neanderthal The elder also showed about 5% Japanese. Not so surprising, being from a southern coastal province. Oddly enough, the ex- and I show none of those traits except the Neanderthal. Of course they both have the markers for progressivism and atheism and the younger tested positive for being a NE Patriots fan (she got that from my side of the family

  5. says

    Making statements about fear of heights and public speaking is in some ways more problematic than statements about one’s eye color or dandruff. The latter are facts one can easily check by looking in the mirror. The former are nebulous statements that can potentially cause confirmation bias: “My daughter’s DNA test said she’s likely to be scared of public speaking, so we should discourage her from any career where public speaking skills are necessary.” Or alternatively: “My DNA test says I’m likely to be afraid of public speaking. This means my nervousness is genetically predetermined, and therefore it’s hopeless for me to overcome said fear, hence I won’t even try.” Statements like these make me think about the false belief that girls are supposed to be bad at mathematics. Thus many girls don’t even try to get good grades in their math classes. Many math teachers expect girls to fail and don’t even try to help them more. The result is a self fulfilling prophecy.

    I really hope that DNA tests won’t start telling people shit like “you are a born leader,” or “you should be a stay-at-home parent.”

  6. nomdeplume says

    Myers-Briggs and astrology are good comparisons. Things like “fear of heights” and “fear of public speaking” and “ability to match musical pitch” are clearly not “genetic”. The rest of it I presume is, though trivial. I don’t need a genetic test to explain my baldness, a look at photographs of my father and both grandfathers does that. All part of the modern misuse of science which is helping to shape public attitudes to science.

  7. hemidactylus says

    My dad was gregarious and capable of making friends with anyone. My mom was reclusive outside of work and her bingo neighbor. I am socially anxious. I could have learned social skills from my dad but didn’t. But my mom and one maternal uncle went gray to white hair and like my dad I have so far held hair color. One maternal uncle went top bald and I haven’t yet. Not sure where the back hair came from. Both parents were conservative. My living maternal aunt and I are both liberal.

    Tempted to do 23andMe to get the scoop on actual heritage. My guess, if not secretly adopted, is half Swede and a mix of Scots-Irish (orange vomit), Brit, and Dutch. Would be cool to get a curve ball from the Levant. Or some other exotic locale.

  8. jrkrideau says

    Myers-Briggs and astrology are good comparisons.
    This is being very unkind to astrology which at least started out with observable astronomical facts, REAL STARS!

  9. nomdeplume says

    @10 Well, when you say “real stars” you mean stars that made the shape of a bull, or an archer, or a fish, or…

  10. dianne says

    I stopped getting tested after doing the National Geographic genographic project, which only gave mitochondrial and (for my father) Y chromosome haplogroup. My mitochondrial DNA is haplogroup A, which is seen in Asians and Americans. The specific mutations look like probably mesoamerican but the tribe is…probably no longer existent. The 10,000 or so person database had two matches for me, specifically my sister and my mother. Well, at least we had an identifiable region of last mutation. Then there’s my father. The Y chromosome is general haplogroup G, which is characteristic of…pretty much no one. It shows up here and there in Eurasia, but with no particular pattern. Then there’s the details. The report used the term “unheard of.” Several times.

    Conclusion: I don’t really know who my people are but whoever they are/were, they are and were absolutely shit at reproduction. On the plus side, I’m almost certainly not related to French or Austrian royalty. (Sorry PZ!)

  11. John Morales says


    I don’t really know who my people are but whoever they are/were, they are and were absolutely shit at reproduction.

    Good enough, but. After all, here you are. ;)

    re the OP:

    For instance, they’ve informed me about traits I might have.

    Would be nice if they could inform about traits one cannot have.

    (But then, the very term ‘trait’ is a bit weaselly)

    Re the OP title:
    “Tribe” has multiple senses; one is about social kinship, another is about biological kinship. The former, one does not need to be told.

    (And, obviously, not all people!)

  12. PaulBC says

    I have siblings who did 23andMe. It backs up my Irish lineage (mostly in the US for a long time, pre-potato-famine, but all from Ireland) so it’s not very interesting. I consider my tribe to be the science fiction fans I hung out with in college decades ago.

  13. wzrd1 says

    A few things.
    First off, per that service, neither of our children are related to myself or my wife.
    Given other independent tests, one individual sample presenter was not related to himself and later, herself, via two ‘parental samples’.
    Now, chimera is a real thing, but the divergence reported was insanely wide, well and away from chimerism.

    I trust these “genetic testing services” maps about as well as I trust racist bullshit, it’s just that bad of a pseudoscience.
    Pity, real science has done a lot better, just not as “granular” as the pseudo-scientists judging bumps on the noggin.
    Any such is welcome to study mine, got head butted by my cousin’s pet goat, the goat sat down with splayed legs, while I exclaimed “Ow!”.
    That, actually, a true story.
    Idiot me bent down to get face to face with the goat, hopefully, to be less threatening via towering loss, it expressed affection and playfulness.

    An hour later, we played tag, I tackled it and tickled it.

    And for the record, goat does taste wonderful. My last two gallons of pasta sauce was goat meat flavored, by a different goat.
    Did goat for a change, usually, I use mixed lamb and pork.

    The relative was rid of said goat once it gave her a literally dinner plate sized bruise on one thigh.
    Out of respect to her elderhood, but also out of return for a claim of abandonment when I was ordered to go to war and she claimed I abandoned my own father, yes, literally, a dinnerplate sized bruise, with plenty of room.
    My own size, a towering 5′ 9″, slightly below average body weight specimen.
    Pity, that goat could’ve brought me plenty of new goats, whatever sex it was.
    I am, after all, an obligate omnivore.

    And yes, I’ve killed an animal and ate it. i strongly recommend all that consume meat do so, to respect that lost life properly and not waste what was gained.
    As to how I don’t have brain damage, never said I didn’t, but I’ve long been proficient at adsorbing a blow, relaxing muscles to move forward.
    Ass hitting ground, goat hitting my head, some towering jedi lookalike slamming a rifle into my head.
    Oddly, all survived the experience, most of the latter were discharged from GITMO.

    Welcome to my odd life.
    Frankly, I’d prefer peace and quiet.

  14. mountainbob says

    I’ve been pronounced less likely than most to sneeze after eating chocolate. So there!