Why DTC genetic testing is good for research

I’ve been reading with interest Daniel’s coverage of the recent FDA hearings into DTC genetic testing. In this context, both he and Razib Khan are incensed by a video which seemingly shows an FDA official misleading Congress about the research done by 23andme:

You can think what you want about the value of the research done to date by 23andme [1], but in my mind, there’s one simple reason why the sorts of participant-driven research they’re doing can only be a good thing: all research is driven by curiosity, and the people most curious about a disease or trait are those who have it. While people may think of the academic research community as a machine with endless resources and limitless motivation, it’s not. People work on things they think are interesting; they sometimes follow “trendy” topics, or move into fields with more grant money, or get bored of a given problem and move on. So if the research in the trait you’re most interested in isn’t moving fast enough for you, well, tough luck.

Recall that one of the key players in the discovery of the gene for Huntington’s disease was a foundation started by a man whose wife had the disease (startlingly, the current president of the foundation apparently accused DTC companies of “raping” the human genome during the present FDA hearing). Recall also that James Lupski, curious about the cause of his Charcot-Marie-Tooth disease, simply sequenced his own genome to find it. These are simply well-connected and trained people driven to find a gene involved in a disease. Patient communities that currently exist are also curious and driven, but in many cases are dealing with complex diseases that are amenable to genetics only with large sample sizes and extensive organization; what these communities can now do is outsource, in a sense, their research to 23andme (see, eg., 23andme’s Parkinson’s study). For scientific knowledge, this can only be a good thing.

[1] To date, the novel associations discovered by 23andme are in hair morphology, freckling, photic sneeze reflex, and “asparagus anosmia”. What these things have in common is that they’re biologically interesting, but not particularly medically interesting; it’s pretty much only curiosity that would drive you to map these traits. Medical researchers tend to scoff at this sort of thing; I think it’s actually pretty cool.

  • Digg
  • StumbleUpon
  • del.icio.us
  • Facebook
  • Twitter
  • Google Bookmarks
  • FriendFeed
  • Reddit

6 Responses to “Why DTC genetic testing is good for research”


  • Linda Avey

    Thanks Daniel for calling attention to one of the key aspects of 23andMe–our research mission–that was the cornerstone of the company’s founding but is often either ignored or patently denied (as, apparently, in Shuren’s case). Human subjects in traditional research models are typically defined only by their disease, or as disease-free controls, which is a very narrow and limited approach. And, as most people realize, disease definitions themselves are often 19th-century-derived symptom-based constructs that are sorely lacking in their specificity (Clay Christensen does a terrific job explaining this). So by enrolling these under-characterized subjects in studies that then try to isolate the genetic underpinnings via GWAS, it’s no wonder that many of these studies fail. The 23andMe approach is admittedly a lot messier…imagine the horror of giving living, breathing participants more of a voice and being allowed to, yes, self report…but definitely worth pursuing. Let’s hope the US government officials we pay to do their jobs don’t let histrionics and scare tactics rule the day.

  • Linda Avey

    Thank you, Joe! :)

  • Daniel MacArthur

    Nicely said, Joe. In the same vein, it’s also worth noting that Stephen Quake (founder of Helicos) had his genome sequenced and analysed partly due to a family history of sudden cardiac death.

    And while the novel associations uncovered to date by 23andMe are not medically interesting, they’re doing a reasonably impressive job of targeted recruitment of disease sufferers. Their cohort of Parkinson’s patients, for instance, would be the envy of many a medical researcher. I’m still not totally convinced that they’ll be able to compete against the big academic consortia in finding common disease variants, but I suspect their participants will be far more amenable to studies of long-term disease progression or longitudinal drug response (for instance).

  • As a recent 23andme customer I think it’s fair to say that this aspect of the the 23andWe community was probably the biggest single factor in my decision to submit a sample for testing. I’m realistic to know that the current state of genomics knowledge can at best give indications of marginal risks for the various health conditions, and the fact that my results show a large set of common alleles (‘common things are common’ is one of the medical doctrines that applies equally well here) adds to the feeling that the ‘traits’ data, and the contribution to a developing field of science is at least as much part of the value of 23andme’s product as the medical report.

    I’m still getting strange looks around the office (and given that my office is a biomedical research unit, that’s probably not indicative of the general population yet) for being quite so excited about the possibilities presented by my data. The FDA hearings seemed a world apart – while they were wailing about the risks and the dangers presented by such an ‘unregulated’ flourishing of DTC genetic testing, I was rejoicing in the possibilities presented by such a large amount of data. Maybe the FDA might like to ask a few more people who’ve actually paid for DTC testing why we’ve done it, before deciding that we need to be protected from our own naivety.

  • Mr. Pickrell,

    1) I am struck by the fact that the M.D. commenter was either engaging in FUD or was inadvertently spreading FUD in the comments of my website. How many people are going to know if these companies are, or aren’t, certified, if someone warns about such a lack?

    2) Call me Razib :-)

  • Harry Banaharis

    I’m surprised nobody mentioned Decode.

    Re: Mr Latta. Perhaps the public should have been protected from the advent of the personal computer also!

Comments are currently closed.

Page optimized by WP Minify WordPress Plugin