Updated on July 28, 2018
23andMe has always planned to sell access to its customers’ DNA—a fact it has not exactly kept secret.
When the company’s DNA-testing service launched in 2007, Wired touted its quest to amass a “treasure trove of data ... to drive research forward” as a “key part of the 23andMe business plan.” Co-founders Anne Wojcicki and Linda Avey outright told the San Francisco Chronicle that selling kits was only the first step. “The long game here is not to make money selling kits, although the kits are essential to get the base level data,” a 23andMe board member said to Fast Company in 2013. “Once you have the data, [the company] does actually become the Google of personalized health care.”
So this week’s announcement that GlaxoSmithKline is investing $300 million in 23andMe and using the DNA company’s de-identified, aggregate customer data for drug research is very much in keeping with the long-term business plan. You don’t make that kind of money selling $99 spit kits.
23andMe customers can opt of out their data being used in research, but the vast majority of its 5 million customers have opted in.
The deal comes at a time as pharmaceutical companies are increasingly looking to DNA for new drug ideas. In 2015, 23andMe announced its first partnership with a pharmaceutical company—in which it would study Parkinson’s with Genentech. The deal was reportedly just the first of ten at the time, according to Forbes. 23andMe has also since published studies with scientists from Pfizer, Janssen, and GlaxoSmithKline.
Elsewhere in the world, the Icelandic biopharmaceutical company deCODE has sequenced over half of Iceland’s adult population to identify genes associated with diseases. deCODE was acquired by Amgen and then the Chinese pharmaceutical company WuXi in 2015. A consortium of six big drug companies—AbbVie, Alnylam, AstraZeneca, Biogen, Pfizer, and Regeneron—are paying to sequence all the genes from 500,000 participants in the U.K. Biobank, in exchange for exclusive access to the data for 12 months. (23andMe uses a technique called genotyping to spot-check 600,000 locations in the genome, but it does also entirely sequence the genes.) Regeneron also partnered with Pennsylvania’s Geisinger Health System to sequence 250,000 of its patients.
The idea is to leverage large datasets to find people with unusual mutations that can point to targets for new drugs. The best example of this is a gene called PCSK9, which scientists in Texas found by chance when studying people with unusually high and unusually low cholesterol. Drug companies jumped at the discovery, developing compounds that can target PCSK9 to lower cholesterol. Although PCSK9 is the poster child for genomic-driven medicine, the resulting drugs have struggled to gain a foothold in the market. Critics say they’re not effective enough to justify their high cost.
The idea that customers are paying to have their DNA sequenced, only to have the drugs developed from that data sold back to them has sparked some backlash against the 23andMe and GlaxoSmithKline deal.
If you have previously opted in to 23andMe research and decide to opt out, your data will not be included in new studies that start more than 30 days later, but it will remain in ongoing studies with drug companies. Earlier this year, Bloomberg reporter Kristen Brown tried to delete her DNA from 23andMe and found it nearly impossible in part because, she wrote, “I’d agreed to contribute my information to research, and the company couldn’t retract it from studies in progress.”
23andMe’s first deal with a drug company in 2015 concerned research on Parkinson’s—a disease that 23andMe has long engaged with the patient community. It launched the largest genetic study of Parkinson’s back in 2009. (Wojcicki’s ex-husband, Google co-founder Sergey Brin, carries a mutation for Parkinson’s and that bit of genetic happenstance drove early research interests at 23andMe.) Parkinson’s patients are also particularly motivated to participate in research.
Since 2015, DNA testing has become far more mainstream. 23andMe has added millions of customers, many of them less interested in the genetics of disease than their genetic ancestry. It’s one thing to build a database of customers interested in the genetics of disease and sell it to drug companies. It may be another to build the database out of people who simply want to know where their ancestors came from.