Farming the Madding Crowd -- 23andMe announces its first patent

The direct-to-consumer DNA genotyping company, 23andMe, announced on Monday that it was soon to receive its first patent, one that it had applied for in 2010 (that patent application is here).  It has to do with genetic polymorphisms associated with Parkinson's disease (PD), and, while we're not patent lawyers by any means, the application seems to cover genetic variants and testing.  This announcement is a surprise to many, including many of its customers (indeed, the consent form says nothing about possible patents), in part because the company has come out against gene patenting in the past.

Ken and I have had issues all along with the scientific merit of what these companies are selling (see this post and this one, e.g.).  And we're by far not the only ones.  Much of what they report back to consumers are things the consumers already know (e.g., hair color, freckles), and the estimates of most disease risk, which they also report, just aren't ready for prime time for all sorts of reasons that we and others have commented amply upon.  So, while it may be fun for people to know their genotypes, and sometimes their ancestry estimates, the information is generally not very useful clinically.  And apparently it's not just for public edification or recreation: it is run by Google money, after all.  We believe it's unethical for any of these companies to be selling disease risk, no matter how many caveats they include, and this is before considering the profit motive in patenting what they have found apparently on the backs of their customers.

How does 23andMe work?  You spit in a tube and send it to the company along with your payment, and they genotype you at selected sites and then give you full access to that data, with some explanation.  If you want to take part in the research the company does then you can choose to answer surveys about lifestyle or other risk factors, which is your consent for them to throw that information and your genetic data into their research pool. The company says that most of their customers have chosen to do this, but at the same time, it seems that none of their customers were told that gene patents might result.

Gene patenting has been controversial since the first one was applied for, in part because it is difficult to know how to apply the generic wording on patents in the Constitution to the molecular age. Basically, patents should cover inventions or discoveries, and while scientists may in fact discover genes, the law now says that they can only be patented if the scientist has also discovered a function for the gene.  That's a law that had to have been written by not-disinterested geneticists!  Legally speaking, the polymorphisms discovered by 23andMe for PD are patentable, but is that really innovation?

But legal or not, many, including us, are still not in favor of gene patenting.  To us, patenting should be not getting to own properties of nature, but for innovating value-added products.  A medical test, that any beginner who works in a genetics lab for a week could replicate, shouldn't be covered.  If the patents were designed to protect against commercialization of naturally occurring genotypes, as NIH once threatened to do, that's a worthy public-domain protectant--against predatory commercial practices--but that's not the idea here.

The gene patenting issue has been endlessly discussed and the discussion is easily accessed online, so we won't repeat the arguments here except to say that our personal concerns are that patenting makes public and often publicly funded information private, for private personal gain.  In addition, it can tie the hands of clinicians who want to be able to offer genetic testing to their patients, it can prohibit others from doing research on a gene of interest and so on, and while we may be a minority we're not in favor of the get rich quick motivation for doing research.  Some "defensive patenting" has been going on (patenting to protect against genetic profiteering), but in our opinion that should not be necessary -- naturally occurring genes should not be patentable, period.  

As if the issue of this 23andMe patent weren't problematic enough, it seems that the company's motivation for focusing on Parkinson's disease is a very personal one.  Apparently the CEO's husband, one of the founders of Google, has been found to have a PD risk allele, and the couple has donated millions of dollars to the Michael J Fox Foundation (which will be one of the beneficiaries of this patent, along with Scripps Research Institute) for research.  Well, fine, but they should have been up front about their personal interest in mining the 23andMe consumer database for PD variants and asked their customers whether they were willing to donate spit, and pay for what turned out to be a PD research 'crowd-funding' project (see our post last week on this idea as related to fairer science funding).  Instead, they appear to have done it surreptitiously, complicating what's already making many 23andMe customers feel misled, if not betrayed (see comments on Twitter for examples--here's a randomly chosen set of them, e.g.). Not to mention that some of the scientists affiliated with this company have had prior apparent interests in ethics, to boot.