23andWho?
Is the personal-genomics bubble bursting? That seems to be the implication of Andrew Pollack's recent feature in the New York Times: "The scarce ingredient so far is customers."
As a consequence, DeCode Genetics has been through bankruptcy and repositioned itself to sell to doctors rather than the general public. Navigenics has changed management and laid off workers. The most high-profile company, 23andMe, is down to about half the workforce it once had, and in contrast to its previous publicity-seeking habits, refused interviews for the New York Times piece (except for a last-minute email).
23andMe, in particular, had hoped to develop a large database of DNA profiles, but even massive price cuts are not attracting the customers: A $25 offer to people suffering from Parkinson's attracted well under half the company's goal of 10,000 takers. And now technological advances may be threatening their entire business plan.
Genomics in general and commercial personal genomics in particular developed on the assumption that common diseases were closely related to common genetic mutations. Therefore, it would be effective to base predictions on analysis of a very small sample of the genome. That assumption now seems to have been wrong. Hence, as Nicholas Wade explained in the New York Times, "a new approach is needed, one based on decoding the entire genome of patients."
That is starting to happen. The cost of sequencing continues to plummet, and academic researchers are beginning to sequence the entire genomes of patients, and even their families, rather than the small portions these commercial companies are analyzing.
Moreover, we reported eighteen months ago the speculation by David Hamilton that:
Once reputable medical institutes get into the genomics-research game in a big way, research scientists aren't going to pay out significant fractions of their grants to get access to a commercial database when noncommercial databases - quite likely of higher quality - are more readily accessible.
So 23andMe's databases may be inadequate, undesirable and ultimately unprofitable. The company seems still to have plenty of cash, and it can perform some limited predictive analysis, but it's looking more and more like a small-niche business. Recreational genomics may be a fad that has passed its peak.
Previously on Biopolitical Times: