The Necrogenomicon

Msgr. Falda: The language of the Necrogenomicon is arcane; its meaning recondite. If we give it, in its many variants, into the hands of the people, there's no telling what they may do with it!

Bro. Laxman: But, with respect: it already resides in their hands, and hearts, and indeed in all parts of them. It lives in them, and through them, and they in and through it. Most literally.

Msgr. Falda: But it requires interpretation. Trusted interpretation. Authoritative interpretation….

Bro. Laxman: To be sure, many details of life benefit from the wisdom and insight of experts. But nobody wants to do away with authorities and experts. It's merely that the people want to read the language of the text themselves, and perhaps consult with others who know more than they.

Msgr. Falda: This cannot be! If the people read for themselves a text they do not, and probably cannot, comprehend– and if they follow the guidance of whomever they will rather than that of a rightful shepherd of the flock– then they may go astray, not only in understanding but most certainly in action as well!

Bro. Laxman: But the people may already consult whomever they will, and go as they choose, and understand according to their lights, and act, possibly, in manners untoward.

Msgr. Falda: Precisely! And uncovering these truths to them all at once, in bulk, and without appropriate commentary may mislead them further! What if one of them comes to a false understanding and seeks to cut off his right hand?

Bro. Laxman: We already govern the chirurgeons, my lord.

Msgr. Falda: But… but what if one of them seeks to foment rebellion?

Bro. Laxman: We already regulate the militia, my lord.

Msgr. Falda: And what if one of them, for want of understanding, annoys a deacon with babble and the ill-gotten fruit of a meandering mind?

Bro. Laxman: Then he will tell him to stop, my lord. And perhaps help him to understand the limits of his own horizon. Knowledge is seldom fatal, and even a false understanding will seldom bring about grievous harm….

Msgr. Falda: But we are the gatekeepers, Bro. Laxman! We are the gatekeepers.

Bro. Laxman: And each of the people, my lord, is the gate. Shall we keep it closed and guarded as for war, or open as for peace, its perimeter defended?

The FDA reckons that the product provided by 23andme is medical equipment, and that some subset of the corresponding service constitutes medical advice. So the FDA wants a piece of the actionto be sure that the people are protected from the dangers of possibly false or misleading information coming through unauthorized, unregulated channels. 23andme has been draggin' its feet in response to FDA demands, perhaps because of disagreement about whether personal genomics, a new application of new technologies, actually falls squarely within the current regulatory regime.

BoingBoing provides a cartoon and a cluster of links to articles that offer a fresh and useful overview of the issues at hand.

A bunch of dead people gave me their chromosomes. Ever since, I've been trying to figure out to how organize and use them. Not too long ago, I sank a Frank' into the "Health and Ancestry" personal genomics kit from 23andme. Just in time, since the FDA has asked them to stop making sriracha until the neighbors' complaints can be mollified. Last I heard, 23andme is making nice in words about compliance and cooperation but declining actually to comply… for now. "Can't we all just get along? I'm sure there has been some sort of misunderstanding. We've made a hash of it with our tardy replies, but we do, genuinely, truly, from the bottoms of our heart, love and respect you. It's not you; it's us."

(BTW, feel free to use me as a referral once they sort things out! That'll add $5 to my book-buyin' fund. ;) )

Did the results of my test solve any deep mysteries? No, although I learned some things about my ancestry that I hadn't previously known and have since confirmed genealogically. Did health information spur me to bum rush the medical staff at my PCP's office and demand that they do X, Y, and Z forthwith? Not at all. Was it entertaining and informative? You betcha! And did it prompt me to try to learn more about genetics, genomics, and gymnastics? Indeed, it did. I was floored by the exercise, which set a high bar, and I wouldn't call my efforts so far a ringing success, but that's ok since I'm just horsin' around.

Herewith, some observations. First, 23andme takes a conservative approach to analysis; if you download your genome info, upload it to GEDMatch, and run some alternate analyses offered as freeware by genetic hobbyists or rogue professors, you may see more– or different– information about haplogroup classifications and ethnic origins. Using a different commercial service, such as FamilyTreeDNA, may likewise provide more granular results. But for 99 clams, 23andme delivers the essential and allows some speculative tweaking to see alternate results. That's good enough for the casual consumer; those on a mission may need more.

Second, the community forum at is fairly primitive. For example, email notification for followed discussion threads is an all-or-nothing affair. Searching is non-existent. Redundant threads occur because there's no fast, non-awkward way to find out whether an appropriate thread already exists.

Third (and this is probably true of all personal genomics communities at present because this industry is larval), the points of light are far outnumbered by the blobs of smog. To phrase it with greater diplomacy, the discussion forum is overrun by understandbly curious and uninformed users whose questions, and whose answers to others' questions, are flat out wrong. In the midst of all that noise, a few valiant and well-informed hobbyists (plus the occasional professional) who have dedicated themselves to the task try to set things right. Sadly, the forum software sees those contributions fade rapidly into undiscoverability.

I trust the quality of discussion will improve there, and elsewhere, as education improves and interested parties take advantage. Indeed, 23andme provides a number of informative introductory videos and simple essays that lay out the basics while identifying some of the limitations and nuances. But reading and watching videos are homework, and nothing guarantees (nor should guarantee in that sort of forum) that everyone who speaks has done that homework.

Do you have some experience with personal genomics services? What was your experience? Did you learn anything surprising or interesting that you'd like to share? What do you think of the policy issues underlying the FDA's attempt to regulate 23andme?

Last 5 posts by David Byron


  1. Delvan says

    I ordered my 23andme kit a few moons ago, and was very happy with it. At their price point, they offer more testing than Family Tree DNA (charges separately for Y chromosome and/or mitochondrial genotyping) or AncestryDNA (likewise). My sister then ordered a kit, and a day or two before the newest FDA letter my 2nd cousins and their dad ordered a kit after I explained how it had impacted my genealogy research.

    The raw data, the book, is essentially the same (at the $99 price point) for AncestryDNA and Family Tree DNA, the latter just leave off a few chapters, but they can also sell you some chapters that 23andme doesn't offer yet. The difference for the FDA is, apparently, because 23andme will give you a lot of pointers on what we think we know about some of those words in the common tongue, while the others don't comment on the medical side so (as best I can tell) the FDA ignores them.

    I agree, the forum setup is abysmal, the comparison tools have problems when it comes to trying to triangulate (e.g. compare multiple cousins against each other to find who are related by the same ancestral line), and unlike Family Tree DNA you can only do "in common with" searches for people who have manually shared their genome with you (and even that, it's hard for folks to find the best tool to do that as it's hidden in a spreadsheet for a tool you wouldn't expect it to be).

    But you get the best bang for your buck in the actual genome data for you, and you can upload your raw data to gedmatch (free), or to family tree DNA (for a reduced fee, $49) to use their tools on your data (and get access to comparing your genome to people in their databases).

    Its really satisfying when you find that common ancestor from many generations ago that contributed the sequence you have in common with some other cousins. I've started adding a little graphic to my tree on ancestors I have genome confirmation of, as well as then adding that cousins lineage to the ancestor as well.

    Oh, and edited to add: My data is on gedmatch, on Family Tree DNA and soon to be on Ancestry as well. The latter has no "upload from third party" option, so I had to buy a kit from them. It's exactly the same kit, only difference is packaging.

  2. David says

    So the question is whether it is a medical device? Determining this is not always straight forward. For example a tongue depressor is a medical device when sold to a hospital but the same piece of wood at the hobby store is not. Google FDA and medical device and you can see the definition. I believe the relevant part is:

    intended for use in the diagnosis of disease or other conditions, or in the cure, mitigation, treatment, or prevention of disease, in man or other animals, or

    So if they can successfully argue that this is not intended as a diagnostic tool, they should be exempt. If not they need to determine the class of the device and whether they need to notify the agency or seek pre marketing authorization. This is all straight forward stuff for people in this industry.

  3. Rhonda Lea Kirk Fries says

    In 2010, 23andMe told me my estimated risk for breast cancer is below average (10.1% vs. 13.5%).

    In 2011, I was diagnosed with breast cancer.

    Further investigation reveals that my raw data contains some risky breast cancer genes not evaluated by 23andMe's report, but even without that, the heritability of breast cancer is between 27-40%.

    23andMe is very clear about the limitations of its report and of genetics, generally, so I was neither surprised nor angry at the lack of congruence between their prediction and my diagnosis.

    I suspect–based on what I see in various FB and other support groups–that I'm in a minority.

    Currently, 23andMe raw data has become a cash cow for a certain population of medical…um…I'm not sure what to call them, but they refer to themselves as "practitioners," although they are surely not geneticists. They have established businesses that collect largish sums of money to evaluate certain SNPs and to "prescribe" an overabundance of supplements to "address" those SNPs. I see that as a problem, especially if the FDA is aware of it, and I'm quite sure it is.

    I don't know how 23andMe could be more clear about the limitations of its genotyping service. The disclaimer is in plain English using very simple words. I guess no one reads it, because a lot of sick people seem to expect that several hundred or fewer SNPS (out of 900k reported, out of 9M known) will fix their every ill.

    On the other hand, I too take a lot of supplements. Because of 23andMe, however, I know that I'm likely a CYP2D6 intermediate metabolizer. The research indicates that I must take care to avoid any supplement that inhibits CYP2D6 so as not to interfere with the action of tamoxifen. My recently-fired oncologist probably should have warned me about that, but she didn't. Without 23andme, I would have slowly and unknowingly shortened my survival.

    Bottom line, the FDA can go pound sand. The spit test is non-invasive, and it tells me about me. I have a right to that information and to whatever hairbrained actions I want to take based on that information. If the FDA succeeds in this stupidity, the next thing you know, PubMed will shut down public access, and we'll again be on the road to doctor deification.

  4. Shane says

    I took the the Strong Inventory Test it was both helpful and useless. Heaven knows I have probably made some bad career choices because of it, but thankfully I am able to integrate and understand my mistakes. I am a big boy now and my life is a process of trial and error with some error producing disastrous results. As a big boy I have learned to prepare for and mitigate for such eventualities. My solutions are unique to my situation and will probably not work for others. Can I just be left to understand my own life and all of its quirks and irregularities, or must I constantly be nannied into a straight jacket for which I can no longer manifest solutions to my problems.

    Thank you FDA for making the world just a little bit safer!

  5. Red Tonic says

    I'm not sure how authoritative this gossip is, but 23andme is financed by Google and the founders are in no way independent in appearance. The theory springing from this is that 23andme is a way for Google to load up on genome information and mine the heck out of it at some point. That might also point to some federal administrative-level politicking w/r/t putting the kibosh on 23andme's business (for now).

  6. Dan says

    I tried it mainly for the genetic history thing; kind of hoping that it would reveal unexpected origins. Sadly there were no surprises, other then the estimated percentage of Neanderthal genes. As for the medical part of it, I thought it was talking about probabilities only, and as such wasn't to be used to diagnosis anything health related. Maybe the Neanderthal part of me just didn't understand. All in all I thought it was a nifty way for me to spend some disposable income.

  7. Rhonda Lea Kirk Fries says



    And I forgot to say in my first comment how grateful I am for your post about this situation. As irritated as I am by the whole thing, the way you wrote it gave me a good reason to laugh.

    @Red Tonic: Seriously?

    23andMe is financed by many entities besides Google.

    The co-founder of 23andMe (Anne Wojcicki) is married to (albeit, estranged from, but still engaged in philanthropic and business ventures with) the co-founder of Google (Sergey Brin), so it would be hard for them to appear to be independent, because they're not–they even have children together. Such a conflict of interest.

    23andMe has made no secret of what it is trying to do with the data. It's a goldmine for research, which is why, for example, Genentech is currently doing its Avastin study with 23andMe data.

    A better rumor might be that the FDA is trying to clear the way for other business interests:

  8. AlphaCentauri says

    If you can test your own drool, you can test someone else's drool. There may be a concern about people doing testing on other people without their knowledge, or about it being done under duress by employers or abusive spouses.

  9. LauraW says

    I bought a test kit a few days ago when rumors of the FDA letter appeared. It just shipped today. Hopefully 23andMe will run the tests for kits they've already shipped out even if they decide to play nice with the FDA. I wonder if this news has caused a nice sales bump for them… :-)

    @Red Tonic: As far as I can tell, that is just the typical "Google is evil" conspiracy theory with no evidence to back it up. Google is an investor in the company, and Sergey's (soon to be ex?) wife is the CEO, but that's it. Google does not have access to 23andMe's data. If it did, there's a pretty good chance I would know about it. Google is not very good at internal secrecy. Disclaimer: I'm a Google engineer, though speaking for myself. And annoyed at this sort of rumor-mongering.

  10. TW says

    The folks over at ScienceBasedMedicine have a good article on this discussing the pro's and cons from a science and medicine plus ethics angle with a hint of regulatory law. The big issues are science is not what average folk think or wish it was any more than law is. Also fun with statistics.

  11. Deathpony says

    I'm kind of mixed on this one. I can see some of the reasons why there is disquiet…the science behind the medical output they are providing could do with a lot more rigor…something that ironically the data they gather will help deliver. Genomics and informatics is the way of the future, but we are kidding ourselves if we say that we actually have a brilliant handle on them just yet. The cost of screening has come down so much faster than anyone anticipated though, so the process is exploding whether we like it or not, as is the data.

    Even if just looking at SNP screening, I had the experience of establishing a research project in plant genetics to screen for SNP's associated with certain traits in our species of interest that gives an idea. The project was aiming to find 500 SNP's; by the time it finished just one year later, for the budget we were able to do 10,000. Our capacity has grown so much faster and our thinking is still catching up.

    In the end, we need to get better at assiting people to understand the outputs, to be able to provide more robust support on occasions, and to deal legally with this sort of information. In particular, the ability of insurance companies to require disclosure and access.

    I must admit some personal bias that leads me to disquiet on a few levels. I carry the gene for Huntington's disease, and the disease has begun its course in me. I didnt even know I had a potential for it, because I was adopted, so I had no family medical history until I suddenly was contacted by my biological mother.

    Its not a straightforward thing, in these circumstances, deciding if you want to know. The process for screening currently involves a lot of counselling, where it is controlled. Thats the upside of control.

    The reality though is through the proliferation of these kind of technologies, in future, people will be able to screen for a heap of quite bad and fatal conditions without necesarily being prepared to deal with the outcomes of what they are doing. I dont know what I would have done getting that result without any warning…but I am pretty sure it would have been ugly.

    Or for the consequences of knowing for their financial health…I could not get insurance, for example, through my superannuation fund without getting legal on their ass and still the outcome is pretty sub optimal. We dont seem to have dealt with the privacy implications at all well yet.

    So, have at it, but lets understand what we are unleashing, and own the consequences.

  12. Stephen H says

    A slightly different perspective is available from

    Basically, much of what 23andme claims is unscientific claptrap. Snake oil, if you will. Societies have governments to protect us (at least in theory), and most people prefer having their drugs vetted rather than just relying on fanciful claims. I say most, as some people believe in ideas such as homoeopathy, chiropractic and acupuncture – none of which are adequately policed.

    /bugbear off chest, time to continue.

    23andme is providing some claims about you that are not entirely scientific (i.e. they have not been proven, and in some cases have been disproven). Consumers of health advice should be able to rely on that advice – and that's why 23andme is being investigated/banned in several countries.

  13. Christoph says

    So, what if your test results revealed that you're suffering from some minor genetic disease, let's say something like fatal familiar insomnia, which means that at some point in your life, your brain will lose the ability to sleep and you'll go insane a few weeks later and die a few months later?

    For most people, 23andmes service provides some mildly interesting information. For others, it might be life-changing.

  14. En Passant says

    Stephen H wrote Nov 28, 2013 @1:15 am:

    23andme is providing some claims about you that are not entirely scientific (i.e. they have not been proven, and in some cases have been disproven). Consumers of health advice should be able to rely on that advice – and that's why 23andme is being investigated/banned in several countries.

    As a citizen of an allegedly free country, I object to my government telling me that I cannot know the raw data about my genes without intervention of a licensed physician. Any citizen should be allowed to know, for example, his height, weight, body fat content, measured IQ (whether it's meaningful or not), and a host of other measurements of himself.

    By prohibiting 23andme from providing citizens information about their genes, the government is prohibiting citizens from knowing simple facts about themselves.

    Likewise, any citizen should be free to study and learn the implications of facts about themselves. Does some expert think my body measurements mean I am at risk of this disease or that? Does some other expert think otherwise?

    Any citizen should be free to learn facts about themselves and seek opinion of what those facts mean, from any person or institution that they choose.

    Some people will choose charlatans. Others will choose knowledgeable experts. The government can tell me that it considers spiritualists and palm-readers poor sources for analysis of facts about myself, but it damned well should not prohibit me from consulting them if I choose.

    The government can tell me that XYZ company's scales give inaccurate weights. But it damned well should not prohibit me from buying inaccurate scales if I choose.

    The only thing that government should be empowered to prohibit in that regard is fraud, the company selling the scales claiming they are accurate if they are not.

    If government only states that it considers 23andme to be providing inaccurate measurements of my DNA, it is welcome to do so. I'll make the decision about who to believe.

    But the government is prohibiting 23andme from providing accurate information. And the government is prohibiting 23andme from telling me its opinion about what that factual information means.

    So, fuck the government.

  15. Munin says

    The only thing that government should be empowered to prohibit in that regard is fraud, the company selling the scales claiming they are accurate if they are not.

    But fraud is what 23andMe are arguably perpetrating. There report give extensive details of your supposed lifetime risk of various illnesses and conditions vs the risk in the general population. The only guidance on accuracy they give with those is a four star scale and they give the highest confidence level to many of those prognoses. They do have disclaimers about the service but these caveats are not represented in the actual results section of the report.

    Plenty of other companies offer genetic testing services without FDA needing to be approved by the FDA but 23andMe purport to be able to interpret the arcane language of the Necrogenomicon for the laity. That's why they are in such hot water.

    The reason why people are saying it's an attack on access to genetic testing in general is because due to their connection with people like Brin, their canny marketing and their status as darling of TED and various eek and self quantification communities they have become analogous with cheap genetic testing as a whole.

  16. Rhonda Lea Kirk Fries says

    Basically, much of what 23andme claims is unscientific claptrap.

    Actually, what 23andMe claims is based on GWAS (genome-wide association studies). It's not in the business of inventing an analysis but of putting together the studies that have been done to date. 23andMe has the potential to offer far more information, but it has chosen to stick with well-verified data for the purpose of creating health reports. Those of us with a deeper interest are therefore compelled to spend days trawling PubMed and similar sites for more information. We also sign up for the Personal Genome Project and set our genotyping file loose on the world through OpenSNP.

    I love ScienceBasedMedicine, but in my opinion, it tends to take skepticism too far. Often times, the studies are in process even as SBM is ruthlessly attempting to debunk a claim for lack of scientific validation. But as we all should know by now, science makes mistakes. Sometimes really big ones. For example, look at the mess it has made of the American diet and the payday it has created for the lawyers who represent patients with serious and even life-threatening drug side-effects.

    23andMe offers interesting personal factoids and a conversational opener with one's doctor. It also provides an impetus to make positive lifestyle changes so as to circumvent one's potential genetic destiny.

    The answer to this is not a ban by the FDA but more education for consumers (perhaps especially on the subject of epigenetics). If we don't want people descending into woo, we have to give them better options and more understanding.

  17. phanmo says

    Does anyone know if 23andme will stop selling kits outside of the United States?
    I am neither resident in the United States nor a citizen, and the country where I currently reside and the country of which I am a citizen have no problems whatsoever with 23andme.

    Something that seems to have been missed by most people who have commented is that apparently the FDA gave numerous opportunities to 23andme to discuss the issue which where ignored by 23andme. I'm not defending the FDA, but 23andme probably could have avoided or at least postponed this if they had engaged in discussion.

  18. the other rob says

    The problem with the arguments in support of the FDA's position is that they necessarily presume good will on its part.

    However as any number of real world cases (such as the Lung Flute – available cheaply OTC in most of the world, regulated as a medical device, prescription only and therefore much more expensive in the USA) demonstrate, the FDA appears to have no difficulty with snake oil as long as its purveyors pay proper tribute.

    Absent the presumption of good will on the part of the FDA, those arguments appear less plausible.

  19. Chris says

    As a citizen of an allegedly free country, I object to my government telling me that I cannot know the raw data about my genes without intervention of a licensed physician.

    The way I read it, the FDA would not have any objection to a company that simply provided the raw data about a person's genes. What they're objecting to in this case is the interpretation of that data that 23andMe provides.

    A company could probably provide genetic testing services without running afoul of the FDA by simply reporting the presence of absence of specific genes without providing any commentary about those genes' implications for a person's health. Whether anyone would actually be interested in genetic testing that requires lots of followup research on their part is another matter.

  20. says

    The only guidance on accuracy they give with those is a four star scale and they give the highest confidence level to many of those prognoses.

    This claim is incorrect. Each page where 23andme offers an interpretation of the data also includes links to, or citations of, research papers that serve as the basis of (or at least epitomize) the way 23andme arrives at that understanding. It's not just a matter of the star scale.

    There's also a good bit of small print explaining the probabilities and confidence intervals, the reasons for preferring certain measurement thresholds, etc., and there's a fair declaration of the reference populations employed in their assessment of autosomal DNA.

  21. OrderoftheQuaff says

    Today's Word of the Day is "rentiership'.

    If you get your own 23andme report and learn that you will die in six months from a horrible genetic disease, you might do something irrational and untoward, and we can't have that. Only a highly trained genetic counselor can bring the essential dispassionate gravitas to the table to tell you something like this in such a way as to keep your head screwed on straight, and he's worth every penny of that $200 billed to your insurance company for that 45 minutes.

    "So, Dr. Funbugger, what does this mean?"

    "It means you're going to die in six months from a horrible genetic disease."

    "Lord have mercy! Whatever shall I do?"

    "I wouldn't sign a smartphone service contract for any longer term than that, and you should forget about applying to law school."

    "Well allright then! It's such a relief to be guided by your superior wisdom!"

    Only the highly-educated, highly-trained members of the genetic counseling guild are competent to perform these services, and if you lack the hard-won arcane erudition under your belt, you aren't even qualified to discuss it.

  22. Deana Holmes says

    I guess my problem with the 23andme testing is that 23andme has been dragging its feet on whether or not the testing actually does what it's supposed to do (make certain genetic identifications). I was ready to put my Benjamin down because a friend of mine had tried it and it confirmed (for example) his ancestry. And there's some stuff in my tangled ancestry I'd like to get a handle on.

    But one of the FDA's complaints is that the process may not do what it's being marketed as saying it can do. In other words, it's not a matter of whether or not we, the consumer, should have access to this information, but, rather, is the information being provided by 23andme *accurate*. For all we know (given that 23andme has been dragging its feet), it could be 21st century genetic snake oil. I'd personally like to see confirmation that 23andme's test results are accurate and I don't believe that's been forthcoming in all this.

    I think there are ways around the "medical device" label, such as changing the marketing to have it reflect more of a parlor game ambiance rather than "wow, this is a way to get an idea of what possible genetic diseases may be lurking out there in your genome." Or go to Orrin Hatch and gets some tips on how to make 23andme into an unregulated food supplement or something.

    At least the FDA hasn't gone into 23andme's facility like a bunch of jackbooted thugs and set up shop, asking for information straight from their files. I worked for a medical device manufacturer back in the mid-1990s and one thing we were all very aware of is that the FDA could swoop down at any time and ask to see everything we had. (I haven't kept up with the legislation since I left that industry, but it wouldn't surprise me to find out that the FDA actually has to give some notice now before doing the jackbooted thug routine.)

  23. En Passant says

    Munin wrote Nov 28, 2013 @6:05 am:

    But fraud is what 23andMe are arguably perpetrating. There report give extensive details of your supposed lifetime risk of various illnesses and conditions vs the risk in the general population. The only guidance on accuracy they give with those is a four star scale and they give the highest confidence level to many of those prognoses. They do have disclaimers about the service but these caveats are not represented in the actual results section of the report.

    Chris wrote Nov 28, 2013 @7:55 am:

    The way I read it, the FDA would not have any objection to a company that simply provided the raw data about a person's genes. What they're objecting to in this case is the interpretation of that data that 23andMe provides.

    Statistical risks are inherently difficult for non-statisticians to understand fully. Perhaps 23andMe should provided each client a copy of Cramér, Wald or Rho's classic texts. Then each client would have sufficient basis at his fingertips to develop at least as full an understanding of the risks represented by his results as any FDA bureaucrat would have.

    I can go to a fortune-telling swami in weird robes, who can say "I see by the lines in your hands that you will encounter suffering and death from the Great Siberian Itch." He will not be prosecuted for fraud for giving his interpretation of my physiology.

    He could be prosecuted for fraud if he offers bogus cures for more money. But not for merely stating his interpretation according to whatever mystical BS he claims to rely on.

    But 23andMe is not offering cures in return for more money. They are offering nothing more than an opinion in exchange for a fee.

    Would the FDA back off if 23andMe stated in bold text "This result is for your amusement only. Do not rely on it for any medical decision."?

    Back in the day, even states which prohibited contraception devices permitted condoms to be sold "for prevention of disease only".

    But I doubt the FDA would back off. The FDA is seeking to prevent citizens from knowing, without intervention of a government-blessed medical priesthood, information about themselves upon which citizens might conceivably base medical decisions.

    The FDA thinks that you, dear citizen, are too stupid and benighted to make your own decisions. They know better than you, and they are authorized to use force to keep you that way. So they choose to do so. Because they can.

  24. says

    If 23 clearly disclaimed that the report is not intended to be a diagnostic, a prognostic, or a predictive marker of disease, and all risk associations reported should be discussed with your health care provider, I think they end the kerfuffle with the FDA.

    If, however, they keep telling the FDA to bug off, or not return their calls, they risk getting shutdown, like it or not (I don't happen to).

  25. Maria says

    I am not on principle opposed to FDA regulation, but in this case it seems pretty clear that 23andme is not diagnosing anything. In addition, they link to the studies that support their conclusions, which is more than most doctors do.

  26. jdh says

    Just ordered a kit. We'll see if it shows up.

    The website was clunky and slow. I wonder if they are getting a lot of traffic right now…

  27. Rhonda Lea Kirk Fries says

    In other words, it's not a matter of whether or not we, the consumer, should have access to this information, but, rather, is the information being provided by 23andme *accurate*.

    Tl;dr: The Illumina chip's error rate is just as good as, if not better than, most approved medical devices on the market today. GWAS studies come complete with p-values that make it possible to determine just how conclusive they are. 23andMe develops its reports by putting the two together. It leaves out much of what is postulated, either because it doesn't test specific SNPs or because confirming studies remain to be done.

    More about the chip and the studies:

    The Illumina chip used by 23andMe has a very low false positive rate. I would find the cite for that, but Google is overwhelmed by this brouhaha, and I don't feel like wading through pages and pages of irrelevant information just to post a single statistic. Really, it's low.

    The interpretative information comes from genetic studies. You can have a look here for a sample:

    There's an artistic interactive diagram here:

    The autobiographical rant:

    What bugs me–and I meant to mention this in my first comment, but I still have a bit of chemobrain, and sometimes it takes me off-track–is the tolerance of false positives and negatives in approved medical testing. If the FDA has no plans to hold medical device manufacturers to an impossible standard, why on earth would it impose such on a $99 dtc genotyping test upon which no real doctor would rely (at least not without clinical evidence and additional testing)?

    I mentioned that 23andMe puts me at lower risk for breast cancer. What I forgot to mention is that my 2008 mammogram found a spot in the exact location where my tumor turned up in 2011. I had a stereotactic biopsy. No cancer. In October 2010, my mammogram was all-clear. Less than 5 months later, my gynecologist found the lump.

    I knew it was cancer, but one diagnostic mammogram and core needle biopsy later, I was pronounced cancer free–or I would have been except that I had a good breast surgeon who knew what she was looking at on the sonogram.

    The excisional biopsy that followed found a 1.2 cm HER2+ tumor.

    The excisional was nearly rescheduled because the lump was no longer palpable after the core needle biopsy. Radiology was kind enough to screw up its schedule by doing an unexpected wire needle localization. Of course, that required another sonogram because the mammography device still couldn't "see" the lump.

    The subsequent mastectomy/axillary lymph node dissection turned up a 1.4 cm tumor on one of my lymph nodes. 1.4 centimeters. In a lymph node.

    Clearly, my cancer had been growing for some time. Indeed, I'd had a bunch of wonky lab test results in the years subsequent to the initial biopsy. My endocrinologist knew I had cancer, but he couldn't figure out where it was. Every diagnostic came back a-okay, and breast cancer was ruled out from the beginning because the mammogram was clear. But we still have the pesky fact that my tumor was precisely located at the site of the original biopsy, and that it was not detected on any subsequent mammogram.

    If all these approved tests were so damned accurate, I'd have been diagnosed at Stage 1 (or maybe even 0) instead of at Stage 2A, and my 5-year survival would be increased by a significant percentage.

    It turns out that mammogram false positives and negatives are alarmingly high. Core needle biopsies too. I'd like to see the FDA focus its nitpicking on life and death technologies instead of wasting time on a service that provides hours of entertainment and a guaranteed way to drive one's physician insane.

  28. LauraW says

    The one I ordered on Monday shipped yesterday (from somewhere in the Bay Area) and is supposed to arrive Friday.

    On the FDA and DEA, I alternate between rabid libertarianism and fear for the breakdown of modern society. :-) My gut reaction is to make all drugs legal and over-the-counter unless you can hurt someone else by taking them. (Some antibiotics, for example.) But that "hurt someone else" is a slippery slope. What if, by making prescription drugs freely available you cause a huge public health crisis that costs billions of dollars to clean up? I dunno…

    On 23andme there's no excuse, though. It's just information, for heaven's sake. If there are people out there stupid enough to do bad things with that information (and I'm sure there are) that's a valid reason to withhold it from the rest of us.

    I had a chat with my doctor on a related subject a few weeks ago: lab results. She told me that for most patients, she's happy to write "CC: patient" on lab requests so that the raw results get sent to the patient as well as her. There are a few hypochondriac patients where she doesn't do it, but other than that, it's harmless and often helpful.

  29. Delvan Neville says

    @AlphaCentauri: I can think of no reasonable way for FDA regulation to prevent that sort of unauthorized testing. Further, the amount of saliva you need to get in the vial will take you a few minutes of dedicated salivating & spitting…I suppose you could eventually get enough from a person drooling in their sleep, but you'd have to hang out there for a while to keep scooping more off.

    And again…you could just use an Ancestry DNA kit (identical to 23andme) or a Family Tree DNA kit (some are mouth swabs, not sure if they have spit vial versions) to get the very same data, but neither is on the chopping block.

  30. Deathpony says


    Interesting rant.

    I guess all I can say is its rarely as simple as you say. I can confidently say that, as someone who actually has had a genetic diagnosis of a terminal disease, it actually helped me and my wife a lot having access to those services.

    Should they be compulsory? Up to the jursidiction how they want to handle it, but your rather simplistic rant stunk of straw man so badly I couldnt leave it alone.

  31. Shane says

    @En Passant

    Statistical risks are inherently difficult for non-statisticians to understand fully. Perhaps 23andMe should provided each client a copy of Cramér, Wald or Rho's classic texts.

    Nope "Fooled by Randomness" by Nassim Taleb :)

    I am still contemplating that if you have a coin flip then at the moment that you know the answer to the flip the odds are no longer 50%. Shrodinger anyone.

  32. says

    is just as good as, if not better than,

    Have I told you lately that I love you?


    The website was clunky and slow. I wonder if they are getting a lot of traffic right now…

    Yes. It has been slammed intermittently ever since the big news broke. Prior to that point, I had never seen lag there.

  33. Shane says

    @Rhonda Lea Kirk Fries

    What bugs me–and I meant to mention this in my first comment, but I still have a bit of chemobrain, and sometimes it takes me off-track …

    Wait, what you need chemo for that?

  34. Shane says


    My gut reaction is to make all drugs legal and over-the-counter unless you can hurt someone else by taking them. (Some antibiotics, for example.) But that "hurt someone else" is a slippery slope. What if, by making prescription drugs freely available you cause a huge public health crisis that costs billions of dollars to clean up? I dunno…

    In the Philippines you can get all sorts of stuff over the counter that people in this country would be appalled at. The only breakdown of society there seems to be the corruption when it comes to their economy, and of course catastrophic events of the hurricane variety. Or when I was there of the ground shaking variety.

  35. Shane says


    Should they be compulsory? Up to the jursidiction how they want to handle it

    This is totally confusing to me, why on earth would you (or anyone for that matter) want your neighbors to decide how you will deal with this information.

  36. Deathpony says


    I'm not, I am saying that each jurisdiction will end up making its own call on that based on its own biases. Personally, my view on the best model is one where the primary care physician at least is in the loop and provides the option of referral to non-directive counselling should the patient want.

    Not to belabour the point, but suicide risk is a factor in diagnoses for some of these diseases, and follow up in the immediate aftermath matters.

    Also, while I detect a note of disbelief in your response, I guess all I can point you to is papers like the following. It gives an idea why there has been concern and the response of the medical profession was one of harm minimisation from the outset of testing being available


    American Journal of Medical Genetics

    Volume 26, Issue 2, pages 259–270, February 1987


    Attitudes of 69 persons at risk for Huntington disease (HD) were obtained by means of semistructured interviews and questionnaires. About 79% of the individuals said that they would use a presymptornatic predictive test if it were available. All believed that the test should be made available even though there was no cure for HD. Nearly 2/3 of subjects would use the test for prenatal diagnosis, and of these 71% would terminate a pregnancy if the fetus was found to carry the HD gene. Most subjects believed that pretest counseling should be mandatory and many said that testing should be withheld from persons who were psychologically unstable or were threatening self-harm. The data suggest that about 2–6% of persons at risk for HD may have severe psychiatric or suicidal responses to a positive outcome of predictive testing. This underscores the need for adequate pretest counseling and the availability of professional and community resources tó deal with the impact of predictive testing on individuals and their relatives.

  37. Erwin says

    …there's the serious condition issue – wherein physician intervention/counseling does make sense. Making it a mandate is questionable, but I can see it as debatable.

    …then there is the risk assessment issue, where even many doctors show very little appreciation for statistics. Giving results as absolute 5 year or lifetime risk works better than risk ratios. …eg…the Harvard professor and practicing doctor who prescribed mastectomies based on mammography results without biopsy.

    …and lastly…there is probably the fda just recognizing that propagating the results of recent research is kind of irresponsible. Publication pressure and selection effects, particularly in breast cancer because of small sample sizes and probably financial issues result in the vast majority of published results being wrong – on the 90% level for new cancer drugs, eg. The problem is that the follow up studies showing that tend to take decades… I personally favor allowing access and letting idiots make poor choices, but the bar for idiot there is set pretty high… I can understand an argument that allowing the public access to recent research results without providing an understanding of exactly how unreliable they are isn't helpful. A 10-8 pvalue really doesn't indicate that a study is right – just that, as the experiment was designed, assuming the researchers, unlike most, followed good statistical practices, the results are not consistent with random chance.
    Bear in mind that, for the biggest health issue of the century (obesity), research results got it exactly wrong (fat bad) and resulted in changes that arguably made the problem worse (low fat/high sugar foods)…


  38. Drewski says

    The meat of this post aside, don't be too critical of their forum. Here, for example, a "fresh" reply requires one to scroll to the top, encouraging users to simply reply to the bottom comment. It makes email notification of replies a crapshoot, which is arguably worse than not having the option.
    Not that the setup here is bad, but you chose to criticize 23andme based on what's arguably the worst feature of your own site.

  39. says

    Our blog post comment threads are definitely worse than 23andme's community forum. But then, our comment threads are not, and do not pretend to be, a community forum.

    Our community forum, in contrast, offers all the features one might expect of an advanced discussion forum. (It's phpbb3, which we both respect and loathe.) However, we do not encourage its use, and probably will not do so until we've traded up.

    What does this mean? It means that in your attempt to show inconsistency, you've compared the discussion forum of a large commercial website to the default behavior of a small WordPress comment section.

  40. OrderoftheQuaff says

    It is a joy to peek at Popehat this Black Friday morning and find myself accused of simplistic strawmanning, thank you Deathpony. In my defense…

    Like you, like all of us, I too face a certain death sentence, I just don't know what it's gonna be yet. Just for today, as they say in AA, I have the strength of character and purpose to avoid suicide. When you say that somebody somewhere is going to commit suicide if they don't get genetic counseling, isn't that strawmanning too? You suggest that the psychological constitution of your neighbor, a total stranger, is so weak that they must be counseled before they can receive information about their body, and such a suggestion carries an onus of producing evidence above and beyond a study in a trade mag calling for the creation of a captive cash cow its members can milk for $200/45 minutes.

    First, you would leave it up to the jurisdiction, which is not acceptable to me when it comes to information about my body. Your second comment used the word "optional" which is very good, but then you said that the primary care physician should be "in the loop"? By this, you really mean that the PCP should be in control of the loop, vested with the option of withholding information from the patient, don't you? I ask you this because the patient always has the option of putting the PCP in the loop after he receives his 23andme results, doesn't he? You obviously don't trust your fellow Americans to do the right thing with information about their bodies, yet you expect them to trust Dr. Funbugger and the rest of the healthcare industry implicitly to tell them all they need to know, and nothing more, about their bodies. Do you see the problem now? It is not too late for you to reject nanny-statism and nanny-healthcare and come join the growing cohort of fully informed, fully empowered, fully responsible Americans where I stand.

  41. the other rob says

    FWIW, the wife and I have just put our money where my mouth is and ordered two kits from 23andMe.

    Their site accepted our order and a pending charge shows on my credit card account, so it seems that the FDA's prohibition is, indeed, not on sales but on "marketing". Of course, this raises the question of what "marketing" means…

  42. Ryan says

    I'm going to refrain from commenting on the legalities concerning whether or not this is a medical device.

    That said, one of my many sins was is a degree in genetics, so I'm always a little nervous when I see products like this advertised as providing information beyond the name and location of a gene or codon sequence and whether or not it appears in other samples and with what kind of frequency. There are a lot of intricacies about human genetics and genomics that are really not well understood, and even in the areas where some mechanism are well-understood, they're only well-understood by a very small number of people. The idea of people using information in these kits as medical advice – or really as anything beyond novelty – is pretty disconcerting to me.

    I'm all for people discovering more about themselves, and I fully support the expansion of knowledge about genetics into the general public consciousness (maybe we can finally get rid of the evolution/creation idiocy?), but I get nervous when people try applying what is already scientifically kind of shaky without any sort of training whatsoever.

    That said, that's more of a general comment because I haven't dabbled in these services myself.

    I WILL say that for anyone seriously interested in this, there are a great many public research tools for use in genetics and genomics online which I would consider infinitely more reliable than any commercial product; you are far better to get raw sequencing data (or SNP maps, or microarray raw results) and run the analysis yourself than pay for commerical-use analysis which runs the risk of giving you some really shaky conclusions.

    I would suggest that anyone who is seriously interested in this field is better to register in a genetics/genomics class at their local university – or audit the course – and have access to their tools for a semester or two than muddle about with commercial stuff.

    That said – and while it's been a few years since I actively studied/worked in the field – everything I ever used was public-domain, no subscriptions required. Most of the tools are indexed by and accessible through the US' PubMed system:

  43. Ryan says

    Since for some reason I don't get an edit button on this computer (must be the extensions in this Firefox install):

    I just looked at the 23andme website, having never done so before. The following claims right from their frontpage make me exceedingly frightened:
    -"Carrier Status" (that can be reliably determined for very, very few human conditions and usually requires some kind of specialized interpretation because of allelic and SNP variants).
    -"Health Risks" (even less precise than carrier status).
    -"Drug response" (this is basically a crapshoot except for a couple very specific examples).

    Then again, IF (I don't see it on the page?) you get their microarray results in raw form for $99, that's actually a pretty decent deal.

  44. Delvan Neville says

    There are 54 inherited conditions that they test for. I am a carrier for one recessive condition. Clicking on that includes an explanation of what they think they know about it (e.g. dominant/recessive/mixed), references to the research they rely upon to make this report (3 papers from 1998-2000 in this case), and this
    "A genetic counselor specializes in helping people understand genetic disorders and genetic test results. Learn more about genetic counseling here."

    You get all that before you get to anything describing what the condition actually *is*.

    And yes, you get access to your raw data as well to do with as you please. If you opt to download it rather than browsing it on the site, you go through some more warnings/advisement of the medical privacy risks you face if someone gets hold of that file.

    When I signed up, they had a demo person option where you could look through the results of a demo person to see exactly what they do and don't tell you and how they tell it, if you're interested in really looking at what they offer.

    As to merely paying for the results alone, half the value of 23andme (for me) is their rather large database for genealogy purposes. Gedmatch is an option if all you had was the raw results, but only a small fraction of 23andme users will have uploaded their data to gedmatch.

  45. Rhonda Lea Kirk Fries says


    Tu quoque–avoid at all cost.

    As a practical matter, if you subscribe to the comment feed, it solves the update problem handily.


    I suppose DeathPony has a point applicable to at least part of the population. I, on the other hand, am a do-it-yourselfer at heart, so your comment both resonated and made me laugh aloud. Nicely done.


    I'm passing fond of you, too. It's rare for me to read blogs, but I adore Popehat, because I always find new learning here. I'm grateful that the knowledge comes packaged succinctly and, most especially, with humor.

  46. En Passant says

    Ryan wrote Nov 29, 2013 @11:37 am:

    Then again, IF (I don't see it on the page?) you get their microarray results in raw form for $99, that's actually a pretty decent deal.

    Your question about raw data is answered in the 23andMe FAQ.

  47. Rhonda Lea Kirk Fries says


    Then again, IF (I don't see it on the page?) you get their microarray results in raw form for $99, that's actually a pretty decent deal.

    You get four columns x 960268 entries (including the no calls):

    # rsid

    Some of us disregard all the warnings about genetic privacy and use this file to loose our genotype on world through OpenSNP. (There are also several laymen's tools for the hit and run study of personal genetics. Apart from the economic issues, this may be part of the reason the FDA is out for 23andMe.)

  48. Ryan says

    I like that they're giving the raw SNP-map data. It was only after my earlier posts that I saw their mapping is entirely SNP-based – which is why they can make it so cheap. More comprehensive microarrays or full-scale sequencing would be significantly more costly (though even those technologies are rapidly entering the realm of personal affordability).

    After reading their website further, I can definitely see where the FDA is coming from in some respects. 23andme may link to some peer-reviewed research, and their results may come with some careful caveats, but the advertising phrases that grab you on their front page are… well, if not misleading then not the entire truth either. It concerns me that a business is advertising the use of SNP data for the purposes of carrier information, health risk, drug compatibility when that information all comes with a great big asterix beside it, and it concerns me that their initial advertising doesn't make that clear. It makes it look like they're providing a medical service of a kind. Naturally, I'll leave the determination on that up to the lawyers/judges :)

    SNP mapping is a fantastic and powerful tool in the genetics arsenal and has made life considerably easier, but there are some cautionary points that come with SNPs too.

    Generally-speaking, I would be far less uneasy about 23andme if they were advertising themselves as an analytical service that provides mapping data – for use in genealogy and basic risk factors (requiring further analysis by a genetic counselor) – rather than in the ways their website advertises at present.

    Nevertheless, what an intriguing way of sparking public interest in genetics and genomics. Really, anything that gets the uninformed public off the GMO/stem cell/"Deep Blue Sea" sharks! paranoia that they currently associate with genetics is an excellent development.

    (And OpenSNP is such a cool project).

  49. Rhonda Lea Kirk Fries says


    I got that. :)

    Women are socialized to return an "atta boy" upon receipt of an "atta girl" (whereas men seem most often to merely nod an acknowledgement or similar).

    Besides which, this is the Thanksgiving weekend, so I'm counting my blessings as I go. I include the opportunity to read your posts in that list, so I decided the reason why was worthy of mention.

  50. princessartemis says

    Well, I finally got off my duff and ordered a kit. Mostly it's a curiosity, and for a $100, not something I'd normally spend for curiosity's sake, but there is one thing I am mighty curious about which makes the price worth it to me: the state of my DTY1 gene.

  51. Dion starfire says

    This dispute triggered some flashbacks to the OJ Simpson trial*: the issues and realities will be rendered meaningless by a court of law.

    In the Simpson trial, it was because the presented evidence had made the issue too murky for a reasonable person (which, presumably, the jury was composed of) to be even reasonably certain of guilt, let alone the 'beyond all reasonable doubt' requirement of a criminal trial.

    In this case, the issues will be ruled meaningless because 23andme didn't follow the proper procedure to deny the FDA's authority. 23andme initially agreed to provide the FDA with the information and forms the FDA had demanded. It's only now, when the FDA is threatening shutdown, that the issues will be brought up. I'm fairly confident the legal ruling will be some long-winded variation of "Too late, so sorry".

    The next company to try this (or even the folks at 23andme under a new corporate identity) can object from the first demand, and may even succeed in that battle.**

    * Except for the whole 'I want to throw my tv out the window so I don't have to hear about this anymore' aspect of it

    ** They'll likely have a bit of an uphill battle since there's already been a precedent set for the FDA's authority.

  52. Rhonda Lea Kirk Fries says


    Your DTY1 gene is now known as TOR1A. If you try to look it up in the 23andMe raw data as DTY1, you'll come up empty. Here are more dystonia genes (this is a link to the site of the HUGO Gene Nomenclature Committee):

    23andMe reports 4 alleles in TOR1A:


    The first one is a 23andMe internal ID number. It currently has no corresponding RSID.

  53. says

    "Just in time, since the FDA has asked them to stop making sriracha until the neighbors' complaints can be mollified"

    EPA, but not even them, EPA investigated but it was the city and a judge who told them to stop, and yes, the odor was not pleasant. Has anyone ever bitched about the odor from a cookie bakery?

  54. princessartemis says

    @Rhonda Lea Kirks Fries

    Thank you very much! My family has a history of dystonia and I've got a nasty case of it, but it's one of those life mysteries that one just has to know, y'know?

  55. Rhonda Lea Kirk Fries says


    I'm sorry to hear that. It must be a real trial.

    When you get your results back, let me know–I have a lot of helpful links that will save you time.

  56. GuestPoster says

    For my part: the regulation is totally reasonable. Folks, as a society, have decided that medicine, medical devices, medical advice all get regulated to some extent. All the FDA has done is recognized that the company is selling a product that very clearly fits into these categories, and asked 'what is your proof that it works?' Given the bulk of literature out there showing the tremendously high error rate for 23andMe, it is really no surprise that the FDA is demanding this fairly simple test, one they require for every other drug, device, or 'best practice' before it enters commercial production.

    This is why snake oil salesmen are far less common these days, and Dr. Wowzer's Cure-All Elixer is now just labeled 'Vodka' at the store.

    Frankly, I think people have the right to know their genome if they want to, and have a way to access it. I think it's a fairly worthless thing to know, but have no intention of stopping people from doing worthless things (though I might well feel differently if people start ACTING on these results in bulk, and wasting medical resources that could be better applied dealing with real problems). But to go to the example that starts the ATL post: the Msgr. FDA isn't so much upset that somebody is handing out the text of the document, so much as upset that that somebody hasn't demonstrated that the typesetting is correct, and that the text being handed out is the one the customers think it is.

  57. Delvan Neville says

    Given the bulk of literature out there showing the tremendously high error rate for 23andMe…

    A bulk of literature? Could you cite just a few of these sources showing a tremendously high error rate for 23andMe?

  58. Rhonda Lea Kirk Fries says

    @Guest Poster

    You might do a search for "23andMe error rate" limited to the date preceding the FDA's hissy fit (because you'll have a hard time finding the information if you don't). You'll learn that 23andMe claims an approximate 1% error rate, but other sources peg it lower. I believe the 1% figure includes no calls–unlike one testing service, 23andMe doesn't guess.

    You might also benefit from a stroll through this post (specifically, paragraphs 4-6):

  59. Eli the Bearded says

    Prof Lior Pachter of Berkeley has some interpretation of the error issue on his blog. Pulling out some of the bolded sentences in his post:

    The bottom line is that even though the error rate for any individual SNP call is very very low (<0.01% error), with a million SNPs being called there is (almost) certainly at least one error somewhere in the genotype.
    But the way people use 23andme is not to look at a single SNP of interest, but rather to scan the results from all SNPs to find out whether there is some genetic variant with large (negative) effect.
    Whereas a comprehensive exam at a doctor’s office might currently constitute a handful of tests– a dozen or a few dozen at most– a 23andme test assessing thousands of SNPs and hundreds of diseases/traits constitutes more diagnostic tests on an individual at one time than have previously been performed in a lifetime.
    From the estimates of error rates provided above, and using the back of an envelope, it stands to reason that about 1/3 of 23andme tested individuals have an error at one of their “interesting” SNPs.

  60. Delvan says

    Well, it's not a diagnostic test. It's at most merely a screening test. The accuracy requirements are not the same, but I can certainly see why the FDA argues for overlord rights having a hand in the pot.

    That 1/3rd of individuals on 23andme have an error somewhere in their genome does not mean any given screening result has a high error rate (given how many SNPs are tested, 1/3rd actually sounds lower than I'd expect). It's only when you multiply those tiny errors across all the simultaneous tests that puts it likely that you have an error somewhere. Local clinic here does many HIV tests, after testing hundreds of people it's very likely someone got an incorrect result. I would not, based on that, then describe that clinic as having a tremendously high error rate. But we're just arguing semantics here at this point.

  61. Delvan Neville says

    For those of you who aren't 23andme customers yourselves, this email just went out to those of us that are.

    Dear 23andMe Customers,

    I'm writing to update you on our conversation with the U.S. Food and Drug Administration and how it impacts you.

    If you are a customer whose kit was purchased before November 22, 2013, your 23andMe experience will not change. You will be able to access both ancestry and health-related information as you always have.

    23andMe has complied with the FDA's directive and stopped offering new consumers access to health-related genetic results while the company moves forward with the agency's regulatory review processes. Be sure to refer to our 23andMe blog for updates.

    We stand behind the data we have generated for customers. Our lab partner adheres to strict quality standards that are part of the Clinical Laboratory Improvement Amendments of 1988 – known as CLIA. These are the same standards used in the majority of other health and disease-related tests.

    You are among the first people in the world to ever get access to their genomes. You are genetic pioneers. Thank you for your ongoing support and we look forward to continuing to serve you.

    Anne Wojcicki
    Co-founder and CEO, 23andMe

  62. princessartemis says

    Yeah, they just sent an e-mail explaining that they will still send the raw data and ancestry info, and offered me a refund because they understand that isn't what I purchased. I'm still going to do it though. I hope I can figure out how to interpret what I get back. I have to know :(

  63. Delvan Neville says

    You can. Check out Promethease for a close approximation to the health data the FDA doesn't want them to tell people. Its powered by what is essentially Wikipedia for SNPs, so it hasn't necessarily seen the same level of quality control, but it's free and open.

    You can also check out literature on dystonia and see what SNPs that have been identified as being associated with it, and look at your own data (either downloading it or using their online raw data browser).

    Everybody together….

    +1 internets.

  64. the other rob says

    Further to the above; as people who ordered their kit after Nov 22nd, this is the email that we received:

    Dear 23andMe Customers,

    Thank you for your recent 23andMe purchase.

    At this time, we have suspended the health-related part of our service to comply with the U.S. Food and Drug Administration’s directive on November 22, 2013 to discontinue new consumer access during our regulatory review process.

    As a result, this means we will not be able to provide you with access to the 23andMe health-related results from your purchase.

    We will be able to provide you with both ancestry-related genetic information as well as your raw genetic data, without 23andMe’s interpretation.

    We sincerely apologize for the inconvenience.

    We understand this is not the product you purchased. If you would like a full refund for your order, we encourage you to click here and follow the on-screen instructions.

    Requesting a refund will inactivate the kits listed below and they may be discarded:

    [Names redacted]

    Please stay tuned to the 23andMe blog for additional information during the regulatory review process.

    The 23andMe Team

    I guess that's it then. Next time we see anything like what this used to be it will cost tens of thousands of dollars and be accompanied by TV commercials explaining how "trained advisers" will get Medicare to pick up the cost.


  65. the other rob says

    Apologies for the obscenity that terminated my above comment.

    Not in my defence: I was a bit angry – that's no excuse.

    In my defence: I grew up in England, where the "C-word" carries much less weight than it does here and the "F-word" is punctuation.