23andMe is one of three companies currently providing chip-based personal genomics assays, which provide information about up to a million sites of common variation throughout the human genome. These companies provide insight into a limited but informative slice of your genetic diversity, as I discussed yesterday, giving you information about variants contributing to the risk of a number of common diseases and other traits.
Since their launch earlier this year, 23andMe and competitor deCODEme have offered their genome scan services for a fraction under $1,000. Now 23andMe has announced a price slash, dropping their product down to $400.
The price reduction comes courtesy of advances in the genotyping technology from 23andMe's provider, Illumina (I wrote about Illumina's recent success back at my old site), which apparently allow the company to offer more genetic markers at a lower price. Details on precisely which markers have been added to the chip are still pretty thin on the ground, except to note that they were "either unavailable or unrecognized as significant markers" when the original platform was designed.
(Update: from the press release, it appears that the total number of markers assayed will remain around the same (550,000), but "will include a broader range of Single Nucleotide Polymorphism (SNP) variations and rare mutations not found on the previous Beadchip, thereby providing more relevant data on published associations, as well as maternal and paternal ancestry." The "rare mutations" statement suggests that the new chip may target changes involved in diseases like cystic fibrosis, while the "maternal and paternal ancestry" refers to increased marker density in mitochondrial DNA and the Y chromosome, respectively. The latter change will certainly be popular with genetic genealogists.)
I imagine some of 23andMe's early adopters are currently somewhat annoyed at the thought that if they'd waited a couple of months they'd be $600 better off. (I see Hsien at Eye on DNA had a similar thought - she asks whether recent 23andMe customers will get a refund off their purchase, like Amazon provided following a price drop for the Kindle.) However, I'm sure 23andMe will provide a discount for existing customers to upgrade to the new service - they promise to send out "information on how to upgrade in the next few weeks".
You can expect to see similar price drops from 23andMe's rivals, deCODEme and Navigenics - given how rapidly the price of genotyping technology is plummetting, it would be hard for them to find an excuse not to follow suit (Navigenics' offering, at $2,500, is already laughably over-priced). A genuine price war can only be good news for consumers.
Of course, it will be difficult for any private company to compete price-wise with the Coriell Institute for Medical Research, which is currently offering personal genome scans free as part of its Personalized Medicine Collaborative. The Coriell admission process is considerably more onerous than signing up online for a 23andMe Spit Kit, requiring physical attendance at a collection centre, and the information that Coriell hands out is much more restricted (Coriell refuses to hand out results for anything but a handful of markers regarded as "potentially medically actionable" - so no predictions of Alzheimer's disease risk, and also nothing as frivolous as ancestry markers). Still, the existence of a zero-cost competitor always makes a market a little edgy, as Andrew Yates recently argued forcefully over at ThinkGene. Interesting times...
People who take these tests in the early stages have to know they are early adopters of a technology under development. This really is at the vanity stage for most users, so its hard to feel irritated on anyone's behalf that the price went down. That's the risk of early buy-in--the price of getting to be on the cutting edge. Just curious if you have any other analysis besides on the cost question?
Knowing the risks of being an early adopter wouldn't prevent me from being annoyed at such a sudden and dramatic price-drop (over 50%) - not so much at the company as at my own bad timing. Still, I imagine 23andMe will provide an incentive for existing customers to upgrade to version 2.
Analysis outside the cost issue is difficult due to the lack of detail on the extra markers added to the new chip. You'll hear more from me once 23andMe releases some information.
However, in general terms my advice to potential personal genomics customers hasn't changed much since back in April. If you're only interested in the medical implications, you're probably better off simply putting your money in a high-interest account and not taking it out again until an affordable high-quality whole genome sequence (and much more information on the health implications of genetic variants) is available in three to five years' time. If, on the other hand, you're keen on genetic ancestry analysis or just the sheer awesomeness of seeing your own DNA, then $400 for a genome scan actually starts to look like a pretty good deal.
As for which company to go for: Navigenics is humorously over-priced (presumably in an attempt to appeal to the notion that expensive = higher quality) and has a complete emphasis on medical applications that plays to the weaknesses of current technology; give them a miss. Between 23andMe and deCODEme, 23andMe has a consistently superior interface and (I think) a far more entertaining approach to data presentation without sacrificing accuracy; on the other hand, deCODEme offers over a million markers compared to less than 600,000 on 23andMe's version 1 chip. If the version 2 chip provides a major improvement on coverage I'd say 23andMe is a clear winner (especially in the price stakes, after today), but I'm sure deCODEme has a surprise or two waiting in the wings.
Is it true that David Bowie is putting his fortune into a start-up along these lines, to be called "Gene Genii"?!
23andMe v2 is still based on the Illumina 550k design, so it will still have less snps than deCODEme. However because 23andMe previously added well chosen custom content, they had more SNPedia snps. At present
There are 3971 snps in SNPedia.
* 1974 of those SNPs are tested by 23andMe
* 1644 of those SNPs are tested by deCODEme
* 839 of those SNPs are tested by Navigenics
I expect the v2 will further extend this lead, which had been steadily dwindling.
Hi, Dan. Congratulations on the move to the new home. Looks great, so I hope it works out well for you.
I sort of think this announcement may be the beginning of the end for commercial personal genomics, at least unless and until someone figures out how to make money solely by selling the genotype tests. In other words, I suspect Andrew Yates is right, although for the wrong reasons, because the threat here is to the data-access business 23andMe is counting on, not to the cost of the test.
Great article - I certainly agree that the long-term goal of the personal genomics companies (monetising their aggregate databases) is starting to look pretty shaky, but I'm not quite as pessimistic as you are about the future of the industry.
I read that @ science lever, deCODEme is A+
And 23Andme is better in IT and the marketing around personal genome sequencing...
So is this true, ? where can I find the differences between 23andme and deCODEme, and Navigenics, is it more like 23andme or deCODEme ?
I am member with 23andme. But I would like to open a second account with deCODEme or navigenics...
Hi Michael - unless you've got a really strong reason to open a second account, I'd just stick with 23andMe. The other two companies don't offer much additional value over the information provided by 23andMe - certainly not enough to justify the expense (especially for Navigenics).
If you do decide to branch out to another company, I'd advise deCODEme over Navigenics - deCODEme gives you more information, and it's substantially cheaper.