In an effort to advance research that uncovers new associations between genes and diseases, direct-to-consumer genetic testing firm 23andMe has launched a new research platform that it will open up to a select group of outside investigators who then can bolster the company's findings with their own.
23andMe this week announced that it will accept applications from a "limited number" of academic researchers who propose new ideas for analyzing the company's data. These projects can seek to use 23andMe's data in a number of ways, including identifying new gene-disease associations, exploring new phenotypes linked to established gene markers, looking for pharmacogenomic associations, or honing in on the frequency of rare mutations in the larger population.
"Our primary objective with the research portal is to speed the discovery process so that new genetic discoveries can advance understanding of human genetics and biology," 23andMe spokesperson Catherine Afarian told PGx Reporter.
"We wanted to find a way to connect … researchers from all over the world with different areas of specialty to our vast amounts of data," Afarian added. "We believe the Research Portal can do this efficiently and the potential for discoveries is really exciting."
The data included in the portal will be gleaned from 23andMe's direct-to-consumer genotyping service, called the Personal Genome Service. Researchers chosen to participate in the beta version of the Research Portal will have access to genotype-phenotype data collected from 23andMe's PGS customers who have consented to submit their data for research or from participants of company-sponsored studies. The data will be de-identified and provided under a protocol approved by an institutional review board.
According to information provided by 23andMe about its Research Portal Beta Program, selected researchers will gain access to genome-wide association analysis summaries for 50 phenotypes. 23andMe will provide information about its genotyping platform, details about the chosen phenotypes the company collects, and demographic characteristics represented in the data.
"The 50 phenotypes that will be included have not yet been selected," Afarian said. "They will be chosen in partnership with organizations participating in the beta development of the Research Portal."
Eventually, 23andMe hopes that the beta effort will grow from 50 phenotypes to hundreds. As this beta project progresses, the firm will add additional features to the portal that will allow researchers to make more "sophisticated queries and analyses."
Only researchers who work for non-commercial public or private institutions can apply to gain access to 23andMe's research portal. 23andMe will grant collaborators non-exclusive access to the aggregate-level data in the portal for six months. The data "cannot be distributed, shared, or sold to third parties," 23andMe states in an online description of the beta project.
If researchers wish to publish their findings, they must do so in open-access scientific journals and follow 23andMe's publication best practices. "Currently those best practices only allow for the publication of summary statistics for up to 10,000 SNPs," the company stated. Additionally, 23andMe will not cover the cost of additional data collection from research participants, if a collaborators' research proposal requires it.
However, the data emerging from these research collaborations will not show up in public databases. Afarian noted that submissions to public databases are not allowed by its IRB-approved protocols.
23andMe's spokesperson highlighted that the company is working to grow public knowledge of gene-disease associations in other ways. "23andMe already favors publishing in open-access journals and regularly posts to pre-print servers," Afarian said. "The company also posts its research findings on its own website … and incorporates findings into the reports that are provided to customers of the PGS."
In a recent paper published in the European Journal of Human Genetics, policy experts urged payors and health regulatory bodies to instate regulations that would require genetic testing companies to share genotype-phenotype interpretation data in public databases. This type of open environment, the paper's authors said, is critical to improving knowledge of how gene variants are linked to diseases, particularly as whole-genome sequencing technologies uncover more and more markers that we know little about (PGx Reporter 11/2/2012).
Likewise, the American College of Medical Genetics and Genomics this week released a position statement urging payors, regulators, and genetic testing providers to work on mechanisms for ensuring that the interpretation of genomic variants be shared in publicly available resources.
Separate from the Research Portal beta project, 23andMe does "hope to make significant contributions to the understanding of variants of unknown significance," Afarian said. "Our work with exomes is also still in the very early stages and we are in the process of understanding how best to manage this data."
Last year, 23andMe launched an exome sequencing pilot project that offered customers a chance to get 50 million DNA bases sequenced for $999. At the American Society of Human Genetics Meeting this week, the company will present a poster on how it is handling some of the data challenges associated with its exome sequencing effort.
Researchers interested in gaining access to 23andMe's data in the Research Portal have until Dec.14 to submit their applications.