23andMe… and my likelihood of going bald

23andMe pairs its database of genotypes with user-reported phenotypes to uncover new targets for drug discovery

23andMe is a personal genetics company that offers two forms of genotype analysis to consumers: Ancestry ($99) and Ancestry + Health ($199) . Consumers simply spit in a vial, mail it to 23andMe for analysis, and receive detailed reports on Ancestry, Carrier Status, Wellness and Traits.

In addition to providing this consumer service, 23andMe also takes the data they receive from consumer genotype analyses and combine it with user-reported phenotype data (from over 80% of its consumers!) to create and capture value with partner research organizations. Analysis of the more-than 320 million phenotype:genotype data points, has led to 45 peer-reviewed studies in scientific journals, and better understanding of … balding.

Last month, the 23andMe research team announced findings from their collaboration with the University of Bonn, where they identified 63 genetic variants associated with balding, 22 of which were novel. This study of 17,000 23andMe customers both helped advance the understanding of the biological mechanisms behind male pattern baldness and the health conditions associated with it. One particularly lucrative value creating result is the identification of potential drug targets for pharmaceutical companies.

The role of genetics in male-pattern baldness

Interestingly, this data analysis differs from its competitors in the pharmaceutical industry by flipping the drug discovery process on its head. Instead of screening millions of drugs against various targets in a shotgun style approach, the 23andMe model uses its vast database to look directly at statistically significant associations of genotype and phenotype pairs to zero-in on potential targets before recommending any actual drug screening for any condition. Additionally, a secondary outcome is identifying novel drug targets for associated conditions and subsequently being able to repurpose existing drugs for new conditions.

Precisely how much 23andMe captures this value is non-public information, but it is clearly significant. Largely through the development and progress of their “therapeutics” research organization, 23andMe was able to attract venture capital investment at a unicorn valuation ($1.1B in Oct 2016).

23andMe started out as a pure consumer play amassing over 1 million customers of its consumer products. Realizing the value of their data, the team shifted toward their current research and analytics focus. The primary challenges they have faced along the way include growing the consumer business through marketing, developing a reliable user-reported phenotype dataset through clever user experience design, and gaining the requisite trust for effective research partnerships with academic and private institutions… all while keeping the FDA happy.

While this is indeed exciting and potentially promising value creation, the sustainability of 23andMe model is uncertain for two main reasons. First, it will be a long and difficult road to actually prove 23andMe can succeed in helping to bring a new, impactful drug to market. In the 12 years, give or take, that it takes for a drug to make it through FDA clinical trials, both belief (and funding) from pharmaceutical partners could easily run out. Second, as the cost of genetic sequencing plummets, market-leading full-genome sequencers like Illumina could offer similar services at comparable prices with data at a much higher genetic resolution. Technically, 23andMe only analyzes genotype data from less than 1% of the genome through a carefully selected and much simpler diagnostic assay.

In any case, there is (distant) hope for the bald.



The CrunchBase Unicorn Leaderboard




Google AlphaGo: How a recreational program will change the world


Rolls Royce : value creation through big data

Student comments on 23andMe… and my likelihood of going bald

  1. Great post, Erik! The company sounds like it’s doing some really interesting work. One question for you: how to they handle privacy and ethical concerns? I feel like that would be my main reservation in the decision making process of whether or not to use the service, especially if the data would be shared with third-parties. Thanks!

Leave a comment