Open threads at the Open Thread tab every Sunday and Wednesday

How To Use 23andMe Irresponsibly

As you might remember, the FDA stomped on 23andMe for using too many irresponsible genetic tests that purported to tell you things about yourself and your health with limited support. They eventually worked out a deal where the FDA allowed 23andMe to continue to operate, but they couldn’t claim to be able to predict personal outcomes from your genes.

That means if we want to use 23andMe irresponsibly, we’ve got to do it ourselves. Luckily I recently figured out how to do this and it is exactly as much fun as you would think.

If you’ve got a 23andMe account, log in, go to your name and picture on the bar on the top, and click on the little inverted triangle to get the drop-down menu. Go to the “Browse Raw Data” option, which will give you the option to go to a gene or an SNP. Now all you have to do is find an SNP you’re interested in (an SNP will look like the letters “rs” followed by a string of numbers) plug it in, and interpret the results.

Your best bet here is SNPedia, a wiki collection of different SNPs. If you want to know, for example, something interesting about your risk of heart disease, you can search “heart disease” and get a list of the most relevant SNPs (in this case, rs2383206, rs10757278, rs2383207, and rs10757274). If you click on the first, you can find on the top right in little colored boxes that someone with (A;A) at this site has normal risk of heart disease, someone with (A;G) 1.4x increased risk, and someone with (G;G) 1.7x increased risk.

In this case my 23andMe results are pretty straightforward – it tells me I am (G;G), which is common enough in white people (see the little colored bars on the left of SNPedia; the CEU bar is Caucasian Europeans). Other times the results require an extra step. For example, SNPedia’s page on rs1801133 offers three choices – (C;C), (C;T), and (T;T), but 23andMe tells me that I have (A;A), which didn’t appear to be an option. The problem here is that 23andMe is giving me the minus strand – if you click to expand your result, it will tell you that (“dbSNP Orientation: minus”). When it gives you the minus strand, you have to manually reverse it to get the plus strand. Remember, A is the reverse of T, and C is the reverse of G. So my (A;A) is their (T;T), and I have 1.5x risk of various cancers.

This doesn’t necessarily bear any relationship to reality, because genetics studies often fail to replicate, and even when they’re right they might only apply to certain populations, and even when they apply to people usually people misinterpret what they mean. That’s part of why the FDA banned 23andMe from doing this, and part of why the word “irresponsible” is right in the title. Even if these SNPs survive the tests of time and replication, they will explain at most a few percent of the variance in complex traits, and any claims otherwise are exaggeration at best and pure hype at worst.

But with that fair warning, here are some of the genes I think are most fun to look up. I cannot disclaimer enough that this is for your own amusement only and unlikely to resemble reality in more than the most tenuous way and if I imply otherwise it is a silly joke.

Rs909525 is linked to the so-called “warrior gene” which I blogged about in the last links roundup. People with the normal four or five repeat version of these gene are less violent than people with the three-repeat version, and people with the two-repeat version are massively overrepresented among violent criminals. See for example this article. Although this SNP isn’t the warrior gene itself, it’s linked to it closely enough to be a good predictor. This is on the X chromosome, so men will only have one copy (I wonder how much of the increased propensity to violence in men this explains). It’s also one of the minus strand ones, so it’ll be the reverse of what SNPedia is telling you. If you’ve got T, you’re normal. If you’ve got C, you’re a “warrior”. I’ve got C, which gives a pretty good upper limit on how much you should trust these SNPs, since I’m about the least violent person you’ll ever meet. But who knows? Maybe I’m just waiting to snap. Post something dumb about race or gender in the open thread one more time, I dare you…

Rs53576 in the OXTR gene is related to the oxytocin receptor, which frequently gets good press as “the cuddle hormone” and “the trust hormone”. Unsurprisingly, the polymorphism is related to emotional warmth, gregariousness versus loneliness, and (intriguingly) ability to pick out conversations in noisy areas. 23andMe reads this one off the plus strand, so your results should directly correspond to SNPedia’s – (G;G) means more empathy and sociability and is present in 50% of the population, anything else means less. I’m (A;G), which I guess explains my generally hateful and misanthropic outlook on life, plus why I can never hear anyone in crowded bars.

Rs4680 is in the COMT gene, which codes for catechol-o-methyltransferase, an enzyme that degrades various chemicals including dopamine. Riffing on the more famous “warrior gene”, somebody with a terrible sense of humor named this one the “worrier gene”. One version seems to produce more anxiety but slightly better memory and attention; the other version seems to produce calm and resiliency but with a little bit worse memory and attention. (A;A) is smart and anxious, (G;G) is dumb and calm, (A;G) is in between. if you check the SNPedia page, you can also find ten zillion studies on which drugs you are slightly more likely to become addicted to. And here’s the 23andMe blog on this polymorphism.

Rs7632287, also in the oxytocin receptor, has been completely proportionally and without any hype declared by the media to be “the divorce gene”. To be fair, this is based on some pretty good Swedish studies finding that women with a certain allele were more often to have reported “marital crisis with the threat of divorce” in the past year (p = 0.003, but the absolute numbers were only 11% of women with one allele vs. 16% of women with the other). This actually sort of checks out, since oxytocin is related to pair bonding. If I’m reading the article right (G;G) is lower divorce risk, (A;A) and (A;G) are higher – but this may only apply to women.

Rs11174811 is in the AVPR1A gene, part of a receptor for a chemical called vasopressin which is very similar to oxytocin. In case you expected men to get away without a divorce gene, this site has been associated with spousal satisfaction in men. Although the paper is extremely cryptic, I think (A;A) or (A;C) means higher spousal satisfaction than (C;C). But if I’m wrong, no problem – another study got the opposite results.

Rs25531 is on the serotonin transporter. Its Overhyped Media Name is “the orchid gene”, on the basis of a theory that children with one allele have higher variance – that is, if they have nice, happy childhoods with plenty of care and support they will bloom to become beautiful orchids, but if they have bad childhoods they will be completely screwed up. The other allele will do moderately well regardless. (T;T) is orchid, (C;C) is moderately fine no matter what. There are rumors going around that 23andMe screwed this one up and nearly everybody is listed as (C;C).

Rs1800955 is in DRD4, a dopamine receptor gene. Its overhyped media name is The Adventure Gene, and supposedly one allele means you’re much more attracted to novelty and adventure. And by “novelty and adventure”, they mean lots and lots of recreational drugs. This one has survived a meta-analytic review. (T;T) is normal, (C;C) is slightly more novelty seeking and prone to drug addiction.

Rs2760118, in a gene producing an obscure enzyme called succinate semialdehyde dehydrogenase, is a nice polymorphism to have. According to this article, it makes you smarter and can be associated with up to fifteen years longer life (warning: impressive result means almost certain failure to replicate). (C;C) or (C;T) means you’re smarter and can expect to live longer; (T;T) better start looking at coffins sooner rather than later.

Rs6311 is not going to let me blame the media for its particular form of hype. The official published scientific paper on it is called “The Secret Ingredient for Social Success of Young Males: A Functional Polymorphism in the 5HT2A Serotonin Receptor Gene”. Boys with (A;A) are less popular than those with (G;G), with (A;G) in between – the effect seems to be partly mediated by rule-breaking behavior, aggression, and number of female friends. Now it kind of looks to me like they’re just taking proxies for popularity here, but maybe that’s just what an (A;A) nerd like me would say. Anyway, at least I have some compensation – the popular (G;G) guys are 3.6x more likely to experience sexual side effects when taking SSRI antidepressants.

Rs6265, known as Val66Met to its friends, is part of the important depression-linked BDNF system. It’s a bit depressing itself, in that it is linked to an ability not to become depressed when subjected to “persistent social defeat”. The majority of whites have (G;G) – the minority with (A;A) or (A;G) are harder to depress, but more introverted and worse at motor skills.

rs41310927 is so cutting-edge it’s not even in SNPedia yet. But these people noticed that a certain version was heavily selected for in certain ethnic groups, especially Chinese, and tried to figure out what those ethnic groups had in common. The answer they came up with was “tonal languages”, so they tested to see if the gene improved ability to detect tones, and sure enough they claimed that in experiments people with a certain allele were better able to distinguish and understand them. Usual caveats apply, but if you want to believe, (G;G) is highest ability to differentiate tones, (A;A) is lowest ability to differentiate tones. (A;G) is in between. Sure enough, I’m (A;A). All you people who tried to teach me Chinese tonology, I FRICKIN’ TOLD YOU ALL OF THE WORDS YOU WERE TELLING ME SOUNDED ALIKE.

This entry was posted in Uncategorized and tagged . Bookmark the permalink.

73 Responses to How To Use 23andMe Irresponsibly

  1. hawkice says:

    I have fairly extreme skepticism about the tonality thing. Tones used in e.g. Mandarin are just as subtle and present in English, albeit somewhat different tones and with different usage. Have trouble with 2nd and 4th tones in Chinese? Listen to someone list N fruits or car names in English. I betcha you can tell (even without a closing ‘, and’) when the list ends, and that’s roughly as subtle as normal speed Mandarin (for those playing along at home, English lists are 2nd 2nd 2nd… 4th tones, or it sounds like you’re going to keep listing things — actually just trying saying “2nd 2nd 2nd 2nd 2nd 2nd 4th” and you’ll hear it). Any normal English speaker can differentiate between “huh?” and the interested and uninterested ‘huh.’, and those are pure tonal differences as well. There may be a gene making it easier to e.g. hear inflection marking sarcasm, but I’d be astonished if learning Chinese was the first time that particular inability came up.

    • Chris says:

      hawkice, how does this fit with the claim that basically everyone who speaks Mandarin has perfect pitch? (10x as prevalent as for non-tonal speakers.)

      • hawkice says:

        First, let’s note “speaking Mandarin” isn’t a gene. This isn’t a mistake you are making, obviously, but it clarifies a potential source of confusion a reader might have trying to mentally interpolate the inferential gap here.

        Consider the analogous “How does your theory that there is no gene that encodes for video game excellence fit with the claim that people who play video games for more than four hours a day catch falling objects faster?” Like, practice with fine motor reflexes, I guess. That’d be how those fit together.

        So… Practice. If you use a tonal language every day, subvocalize in it every day, yeah, I mean, I’m just speculating, but more common perfect pitch sounds pretty reasonable.

        It’s also worth noting that there’s a lot more precision in terms of the timing of pitches in Mandarin, as well as the fact that English uses tones to convey meaning maybe 1% as often as Mandarin or less. So, while fluent English requires understanding the tonal elements, it doesn’t mean you get as much useful practice by using English.

        • Chris says:

          Right, I wasn’t talking about genetics, just objecting to your claim that tones in English are “just as subtle and present” as in Mandarin. If this claim were true, speaking English would be as useful as speaking Mandarin as “practice” towards perfect pitch. But it demonstrably isn’t.

          • hawkice says:

            A better wording of my point would be that hearing and understanding the difference in tones is just as mandatory for fluent English speech. That’s what I meant by “just as present” — not that the precision or frequency is the same.

            [Edit: I implicitly used the warrant that precision and frequency determine the value of practice. Worth saying that explicitly, because it’s important and falsifiable.]

      • Wulfrickson says:

        Perfect pitch is pretty strongly linked to starting musical training at a young age. I would imagine that growing up speaking a tonal language would have a similar effect.

        Some quick Googling for “perfect pitch genetics” turns up this 1990 New York Times article reporting on a study that claimed the ability was determined by a single dominant autosomal allele, this report from a random Long Island research hospital claiming to have located it to chromosomes 2 and 6, and this UCSF study claiming to have found suggestive regions on chromosomes 7, 8, and 9 (both in people of European and of East Asian descent but not, oddly, in Ashkenazi Jews). It’s a mess, in other words.

        Anecdata: I have perfect pitch; I’ve also played piano since about age 3, and am getting a minor in music at college. My sister doesn’t have perfect pitch, and she also started musical training quite young (though was never as devoted to it as I was). My mother is a very casual musician (she played some wind instrument in high school and gave it up afterward, if I remember correctly), and my father is one of the most tone-deaf people I know.

    • Tracy W says:

      Stupid question time: can’t I tell when the list ends by the fact that there are no more items after it?

      I’ve said 2nd, 2nd, 2nd, 2nd, 2nd, 4th, and the obvious difference there is that my tongue moves forward to compete the 4th. (I have a slight speech disability and was explicitly taught how to make some common English sounds when I was a child, including “th”, so it’s possible I’m just totally distracted by my conscious knowledge of what my muscles have to do to make those sounds.)

      • ckp says:

        You can tell when the list is about to end as the person is saying the final item, by the tone of the final item.

        I sometimes mess with people by switching the tones around while reading out a list (e.g. a phone number) – it’s like watching someone stumble on the last step of a staircase.

        • Tracy W says:

          It strikes me that a list-ending tone is easier to learn than a tone that changes the meaning of a word within a sentence.

          Also, tones that are just used to convey emotional meaning strike me as easier to learn than tones that convey emotional meaning + semantic meaning.

          • Creutzer says:

            On the other hand, that’s just about what one would expect from an adult whose native language has no lexical tones but, of course, sentence prosody – isn’t it?

    • vV_Vv says:

      Did European languages used to be less tonal than they are now?

      Some European languages, including English, mark questions by inverting the subject and the modal verb, especially in formal speech. This is somewhat redundant given that questions also have a characteristic intonation pattern, which is sometimes considered sufficient in informal speech. The “huh?/huh.” distinction you mention is an example of this phenomenon.
      Some other European languages mark questions just by intonation.

      • In all likelihood, IE languages used to be more tonal than they are now.

        Ancient Greek had a tonal accent system with four distinct tones. This is almost the only ancient IE language about which we can tell much about the accent system, but it is considered likely that other ancient IE languages also used tonal accents, and PIE itself is usually reconstructed with a tonal accent.

        Tonal accent systems aren’t quite the same thing as the contour tones used in Sinitic languages, but in any case they’re in no way less tonal than modern European languages.

        • vV_Vv says:


        • Creutzer says:

          This is almost the only ancient IE language about which we can tell much about the accent system

          Well, we know plenty about Vedic/Sanskrit tonal accents because ancient Indian grammarians wrote about it, and we can reconstruct a lot in the Balto-Slavic branch, where it played an important role for a long time (in fact, Lithuanian still has tonal accents to this day). I’m sure there’s also good reasons why Latin having a pitch accent, but no tones, is the standard story, although I’m myself not familiar with them. It also seems quite clear that Latin did not have tonal accents, because surely some Latin grammarian would have remarked something to the effect if there had been two or more different kinds of accents (although as far as I’m aware, we don’t know whether Latin had a pitch accent or a dynamic accent).

          • Thanks. I wasn’t sure about Sanskrit, so I assumed that it had a stress accent. I was vaguely aware of Lithuanian, but didn’t know how much it reflected an ancient Balto-Slavic system.

            And of course there’s the well-known Norwegian pitch accent, though I believe that the Norwegian system is considered to be innovative and that Proto-Germanic had a stress system.

      • Matthew says:

        Some European languages, including English, mark questions by inverting the subject and the modal verb, especially in formal speech.

        Contemporary US English has strongly diverged from British English on this. Even in the most formal contexts, an American will say “Do you have any X?”, not “Have you any X?”

        • call_me_aka says:

          We (Americans) say, “Have you no shame?” and “Need this always be so stressful?” sometimes. But seriously, do Brits actually say that?

          • Matthew says:

            I’ve heard “Have you no shame” as a fixed expression, but I have never heard “Need this always be so stressful” and it would indeed sound jarring.

            I’ve heard the English use the older inverted word order. Not all the time, but definitely more often than we do in the US.

        • Wulfrickson says:

          Frequencies of “do you have any” and “have you no” for American and British English from the Google Books corpus both show broadly similar trends, with a switch toward do-forms beginning circa 1940. The current “do you have any”:”have you no” percentages ratio is 9:1 in American English and 4:1 in British English, though this is comparing tetragram and trigram frequencies and I’d need to dig through the raw data to get the actual counts for a meaningful comparison, so don’t take this as an absolute measure of anything.

          Also, most common words following “have you no” in American and British English. The set phrase “have you no shame” seems to account for a lot of the uses of the phrase in American English.

          • Matthew says:

            “Have you no…” has a clear implied moral connotation. You’d expect it to be used only in certain contexts. “Have you any” does not have the analogous assumption that you ought to have some.

          • Wulfrickson says:

            True nowadays, though I’m skeptical that it’s always been so. Incidentally, comparing “do you have any” with “have you any” gives a very similar historical trend to the comparison with “have you no.”

        • Creutzer says:

          This instance of have is not a modal or temporal auxiliary, though. I don’t know that “Have you any X” is commonly used in contemporary British English, either. But as far as I know, inversion of possessive have was retained relatively longer than inversion of (non-auxiliary) verbs in general – if the literary testimony of Evelyn Waugh is to be believed, at least into the twenties in upper class speech.

    • Bryan Hann says:

      Perhaps the linguistic concept of “minimal pair” is important here.

  2. Kiya says:

    Looks like you mistyped (G;G) for (A;G) somewhere in the paragraph on Rs6265? (The second instance, looking at the link.)

  3. Anonymous says:

    And here’s the 23andMe blog on this polymorphism.

    Did you seriously just use the word “blog” to refer to a single blog post?


  4. This is fantastic timing, because I just sent in my drool sample a couple days ago! (Right after I figured out how thoroughly the raw data can be used to subvert the “no health stuff” rule.) I can’t wait to confirm all the things I already think about myself!

    I think I’m a worrier orchid who can’t understand people in bars, and I’m kind of curious how I’ll feel about it if my genes don’t actually match that.

  5. Derek Lorian says:

    If you download the raw data and get the SNPTips add-on for Firefox, and then point the add-on to the data file, all rs-coded SNPs on all websites will become links to SNPedia and, if you hover over them, will tell you your genotype for that SNP if available. It’s quite handy compared to browsing for each one in 23andMe’s system. Then, if you get someone who has 23andMe health reports to share their genome with you, you can get much of the effect of having health reports with only a small amount of reinterpretation necessary.

  6. James Miller says:

    I made an appointment (today actually) with a cardiologist to get tested for heart disease in large part because I’m approaching 50 and have

    Rs1333049 C,C and
    Rs10757278 g,g

    I have no non-genetic risk factors. Am I being irresponsible? Please be honest.

    • Anonymous says:

      Everyone is at risk for heart disease. What difference would a 1.7x RR make to your actions?

      • James Miller says:

        My perception of the increased risk has already caused me to make a doctor’s appointment. Because of this risk I worry more about heart disease than cancer and as a result drink more alcohol than I otherwise would. If I find from testing that I am indeed at a very high risk of heart disease I will modify my diet and exercise routine and place a high priority on reducing stress.

        • Anonymous says:

          If your diet and exercise were based on a cost-benefit calculation, you could easily incorporate new information. But probably the changes from your current diet to the optimal one are much larger than the changes to the optimal in light of new information.

          In particular, I completely reject that this has any impact on optimal alcohol consumption.

          • James Miller says:

            “But probably the changes from your current diet to the optimal one are much larger than the changes to the optimal in light of new information.”

            This would have been true for the me of a few years ago. But now I have gotten into optimizing my diet to “live long enough to live forever”.

          • Anonymous says:

            I pointed out a substantive disagreement and instead of trying to learn from it, you ignored it and spouted platitudes.

          • Anonymous says:

            That seems unnecessarily rude, Anonymous. James addressed what looked like the main point of your comment. Your statement regarding alcohol was devoid of any supporting or explanatory detail; if you wanted to be sure he would take it to heart, surely you could have given your reasoning or at least made some claim to authority.

          • Anonymous says:

            The point was not to claim authority, but to make concrete claims to identify the source of disagreement.

        • vV_Vv says:

          Because of this risk I worry more about heart disease than cancer and as a result drink more alcohol than I otherwise would.

          Isn’t this exactly the kind of irresponsible use that the FDA intended to prevent?
          Assuming that the evidence linking these alleles to health risk is strong (I know pretty much nothing of genetics, the SNPedia pages you linked cite lots of studies but also say “Insufficiently evaluated pathogenic”), what is the chance that 23andMe made a measurement error?

          I suppose that there is a tradeoff between accuracy and cost. 23andMe is a private for-profit company, which has therefore a natural incentive to reduce costs, and if they make a mistake the chances that you will find out are virtually zero. I’m no economist, but this looks the perfect premise for a lemon market (like cryonics, by the way).

          Weak evidence is still evidence from a Bayesian point of view, but deciding to significantly change your behavior because of it is probably not very rational in this case. The risk ratios reported by SNPedia aren’t particularly large, and once you factor in a realistic estimate of measurement error, they’ll become even smaller.

    • Scott Alexander says:

      I don’t think the genetics should be your big concern here. If I were you, I’d go to your PCP and get the normal tests (EKG, cholesterol levels, etc) done, then do whatever your PCP suggests with the results. Cardiologists aren’t going to have some special test that solves everything – or insofar as they do have very serious invasive tests, they’re not going to do them on you just because you come in and are worried.

      Concern about your heart health at age 50 is always a good idea, but I think it’s worth starting with your PCP and going from there. If she says you need a cardiologist, you need a cardiologist.

      [This is not medical advice]

      When you say you did this today, do you mean before you saw this post, by coincidence?

  7. Sniffnoy says:

    Link regarding “warrior gene” is broken, seemingly due to messed-up quotes.

  8. The link to “this article” in the paragraph on Rs909525 is broken.

  9. Erik Bjäreholt says:

    That last sentence sound kinda warrior-ish.

    Also “1.4x increased risk” could be mistaken for 2.4 times the risk. But maybe that’s what you meant.

  10. Army1987 says:

    If you click on the first, you can find on the top right in little colored boxes that someone with (A;A) at this site has normal risk of heart disease, someone with (A;G) 1.4x increased risk, and someone with (G;G) 1.7x increased risk.

    Is the G allele very very rare, or are they using “normal” to mean something other than ‘average’? Otherwise this violates conservation of expected evidence.

  11. Ptoliporthos says:

    This is on the X chromosome, so men will only have one copy (I wonder how much of the increased propensity to violence in men this explains).

    Absolutely none. MAOA is subject to random X inactivation in females, so both sexes only have one active gene copy.

    MAOB, a duplicate of MAOA is further down the X and escapes X inactivation in females. But nobody ever talks about it being associated with the warrior phenotype, so I always assumed the warrior thing was bunk.

    Now, on the other hand, XX females and XXY males both have a 3-gold higher risk of migraine than XY males. MAOB might be a good gene to bet on there.

    • Hainish says:

      This is on the X chromosome, so men will only have one copy (I wonder how much of the increased propensity to violence in men this explains).

      I was about to point out X-inactivation, but then I realized that Scott may have been making a more subtle point: That since males have only one allele (they’re hemizygous), they can only have _either_ the C or T polymorphism, but not both, therefore no heterozygous men, therefore the effects associated with C are over-represented in men. (This is the mechanism by which men show more variance than women on various measures of interest.)

      • Adele_L says:

        It seems that random X-inactivation would mean only one copy would be active in women also, regardless of their genotype. Unless the random X-inactivation isn’t actually random – which seems unlikely but not implausible.

        • victoria says:

          My understanding is that most loci are inactivated randomly, but some are inactivated nonrandomly. (And I recently learned that a few genes that have counterparts on the Y chromosome are not inactivated at all.)

        • Anonymous says:

          Only one copy is active in each cell, but each copy is active in some cells. The choice of copy in adjacent cells is not independent. The patches of color on calico cats are due to X-inactivation and give a sense of the scale, though the scale varies from organ to organ.

          For example, 1% of men have a defective red cones, so they are color blind. The 2% of women that carry this gene a random half of their red cones defective, but the good half is plenty for color vision. Only 0.04% are as color-blind as men. Thus men have more variance for color blindness.

          More typical is that a heterozygote AB woman would be somewhere between an AA woman and a BB woman, who would be like A and B men. This is also a form of decreased variance.

          Actually, X-inactivation can also increase variance. If 99% of the color genes were defective, then 1% of men would have color vision, while 2% of women would. And while calico can be interpreted as decreased variance in overall shade, they can also be interpreted as increased variance in pattern.

        • Hainish says:

          It seems that random X-inactivation would mean only one copy would be active in women also, regardless of their genotype.

          It does…but if the woman is heterozygous, a random half of her cells will express one allele, and the other half will express the other allele. (This is assuming random inactivation…see the article victoria linked, it has a very good explanation.)

      • vV_Vv says:

        (This is the mechanism by which men show more variance than women on various measures of interest.)

        I thought the standard (just-so?) story for the increased variance involved the fact that some loci escape X inactivation, enabling some degree of compensation within each cell, and that X inactivation occurs in patchwork patterns, enabling some further compensation between different cells. In males compensation between X-linked and Y-linked alleles can occur only at the loci where the X and Y chromosomes are homologous, which are few.

  12. vV_Vv says:

    someone with (A;A) at this site has normal risk of heart disease, someone with (A;G) 1.4x increased risk, and someone with (G;G) 1.7x increased risk

    What do these number mean? Annual probability ratios?

  13. Hainish says:

    This is SO what I want for Christmas, though. To use 23-and-me irresponsibly as heck.

  14. Sarah says:

    I don’t believe in most of these but…yeah I have COMT and MAO mutations, and hella anxiety, and I’m very sensitive to dopaminergic things. Those seem pretty credible.

    Also I have ALL the Crohn’s genes but no Crohn’s or family history thereof. Hope my luck holds.

  15. Alexander Stanislaw says:

    Are you sure you’re non violent? You’re pretty good at beating people up with words.

    In all seriousness, thanks for this, I am now using 23andme very gleefully and irresponsibly. I also have the warrior gene. I have never committed an act of physical violence (or come close to doing so), but when I do get angry (which happens once a year or so), it does shock me. It’s as though there is some blinding force inside me. I don’t know if this happens to other people.

  16. pwyll says:

    If like me you’re lazy and want to get all of this manual SNP checking done in automated fashion, check out Promethease Online, where you can upload your 23andme data and have it automatically go through SNPedia, compare your genome, and give summary and drill-down results. Cost is $5.

  17. mayleaf says:

    Amusingly enough, I was just distracting myself at work by doing *this exact thing* when I decided to check your blog and found that you had posted an article with suggestions about how to do the thing. (I work at a 23&me-like company, we get free genotyping as a job perk so we can be user-testers of our own platform, and I recently got access to my raw genotyping file and was looking up my SNPs on SNPedia).

    Apparently I’m a rs4680(A;A) mutant, which might explain why even low doses of amphetamine-class stimulants have a pretty strong effect on me. I’ve also found that I can get more easily stressed / be less emotionally stable on days I take stimulants. Unfortunately, I also have a hard time focusing at work without stimulants (and am more forgetful, have generally worse executive functioning, etc)… so I’m not really sure what to do about that. Do you have any advice about this, Scott?

    And thanks for the post! Now I have more ways of doing fun, highly-speculative, irresponsible personality analysis with my genetic data 🙂

  18. Douglas Knight says:

    I object to the claim that any of accuracy, replication, simplification, and misinterpretation by laymen had anything to do with the FDA’s decision.

  19. Alex says:

    This reads like high-tech horoscopes to me. I know a couple of friends who took the 23andMe test, so we will have an snpedia party soon. (thanks for the pointer)

    • Nornagest says:

      It’s a high-tech horoscope in that there are an awful lot of SNPs and you’ll almost always be able to cherry-pick a set of them that match your preferred self-image. (Looking exclusively at individual SNPs exacerbates this problem, but other approaches don’t yet eliminate it; I got in to 23andMe before the FDA neutered it, for example, and while they took a decent stab at aggregating data their results still had a strong whiff of this.)

      It’s not a high-tech horoscope in that it’s grounded in something, even if that something will rarely completely determine your personality.

  20. J. Quinton says:

    A while ago I tried signing up for 23andme but on their web form it says they don’t ship to Maryland. Is there a way around this? If I have the sample shipped to a friend in another state and mailed it back from that same state would that work?

    • Anonymous says:

      Here is a comment from someone whose order was canceled because he let them know that he is from Maryland. In that thread, the blogger claims that they pay attention to address, phone number, credit card billing address, and postmark. But I doubt that they pay attention to where your computer is located.

  21. Daniel H says:

    I’m curious why you say it’s irresponsible to use 23andMe like this, but feel that 23andMe was not being irresponsible when they did this themselves. Did they focus on the SNPs with higher evidence, while that’s more difficult manually, or was it something else?

    • Anonymous says:

      I think it is sarcastic. He is letting the FDA define “responsible” and not actually endorsing the usage. Surely Scott’s curated list of 11 SNPs is more responsible than 23andMe’s deluge of information. Most of them are personality predictors, no more dangerous than a horoscope.

      Really, he is telling us how to use it at all, which is a prerequisite to using it responsibly or irresponsibly. If reaching Hope means opening Pandora’s box, so be it.

      Yes, 23andMe’s analysis was more sober than the unfiltered deluge of SNPedia. At the very least, it gathered together all of the SNPs with related effects, rather than letting you fixate on a single SNP. And, yes, they did take into account sample size in which studies to include.

      • Daniel H says:

        I considered that it might be sarcastic, but I decided it seemed unlikely. Given the way he was emphasizing it, I read it as sincere. Maybe that’s not as intended.

        • Anonymous says:

          Maybe he’s sincere. My second paragraph gave an interpretation: he thinks it right to open Pandora’s box, either because he weighs the consequences, or because he favors freedom.

          In any event, the first sentence of the post also calls 23andMe irresponsible, so I reject the premise of your original question. He is merely restoring the status quo.

          • Daniel H says:

            The first sentence says the FDA calls 23andMe irresponsible. It links to Scott saying that it isn’t.

  22. Great stuff. I am extreme on almost all accounts above.

    oxytocin, non-cuddler: AA
    warrior: C
    worrier; smart and anxious (=permabear investor): AA
    divorce, low probability: GG
    longevity; live long and prosper: CC
    popularity: CC

  23. heiner says:

    If you are more computer-savvy, you can also try to use my (free) tool to sample SNPedia, here:

  24. Wisconsin Anarchist says:

    Aside from the “warrior gene,” do you have the SNPs associated with any other genes associated with being a psychopath? I”m pretty sure that my mother is a high-functioning psychopath, and she has given me access to her genome, so I’d like to see what might be there.