Academic Research Collaborations Fall 2017

In addition to providing direct-to-consumer genetic testing through the 23andMe® Service, 23andMe endeavors to advance biomedical science through genetic research.

Some of those discoveries have and will come from our own scientists, but others will come from collaborations with academic researchers external to 23andMe through the 23andMe Academic Research Collaboration Program.

The 23andMe Academic Research Collaboration Program allows academic researchers to access de-identified, aggregated data from the 23andMe database.

The 23andMe database is a rich resource, with genotypic and phenotypic information from more than 1 million of our customers, 80 percent of whom consent to participate in 23andMe Research. By consenting to participate in 23andMe Research, our customers agree to make their de-identified genetic data available for study in aggregate and take part in online research under a protocol approved by an external institutional review board (IRB). These research participants also answer online survey questions on a variety of topics, and to date have contributed over 300 million phenotypic data points.

Recent Collaboration Highlights

Over the years, we have participated in more than 70 collaborations with academic institutions on a wide variety of topics. Here are select examples of projects we are currently pursuing or have recently completed:

  • In collaboration with Prof. Peter Visscher at the University of Queensland, 23andMe is studying the role of genes and the environment on mate choice in humans. Assortative mating occurs when individuals show a preference for those who are similar to themselves. Using data from 23andMe, Visscher’s group has developed methods to understand how much genetics explain similarities between spouses.
  • We have investigated the genetic underpinnings of the onset of puberty in both boys and girls as part of a collaboration with Dr. John Perry at the University of Cambridge. In one study, we investigated voice-breaking in boys and in another, 23andMe results were compared with other datasets to understand age at menarche in girls. We found that shared genetic factors influence puberty timing in both sexes.
  • 23andMe data is also being used to replicate findings, as in the case of a study 23andMe performed to replicate research findings from Kaiser Permanente and UCSF. Scientists at Kaiser and UCSF performed a genome wide association study to identify genetic risks for hernias using electronic health records from 110,000 patients who are part of Kaiser Permanente’s Genetic Epidemiology Research in Adult Health and Aging cohort.
  • 23andMe has been collaborating with medical geneticist Dr. Robert Green and Dr. Scott Roberts, who jointly lead the Impact of Personal Genomics (PGen) Study. The study broadly aims to understand consumer experiences of personal genetic testing — including its impact on health outcomes and whether the testing results in undue anxiety. In a series of papers, the PGen group has investigated the medical, behavioral, and economic outcomes of direct-to-consumer genetic testing, including that of 23andMe customers, in particular.

Our research team has published or collaborated on more than 50 scientific papers. For a complete list, visit

Application Process

We accept applications from academic researchers and twice a year hold a scientific review to evaluate proposals for the limited number of collaborative projects we initiate annually. By opening this opportunity to the greater scientific community, we hope to widen our perspective, spark new innovation, and generate meaningful discoveries from the 23andMe database. In particular, we are interested in collaborations that have the potential to catalyze novel findings beyond what we can accomplish on our own. This includes analyses with collaborators willing to share complementary data resources, new methodologies, or approaches that validate our findings.

What data can be found in the 23andMe database?

We have collected genotypic and phenotypic data from more than 1,000,000 research participants, a cohort that continues to grow. Phenotypic data is self-reported by 23andMe research participants who have responded to surveys hosted on the 23andMe website. A supplement with an overview of the types of phenotype data we collect is available upon request.

What type of collaboration is of interest to 23andMe?

We seek to identify collaborators who will enable new discovery from our database by providing complementary datasets and analytic techniques. We are open to a range of possibilities. Here are examples of the types of collaborations that are of interest:

  • Identification of rare variants with whole-genome sequencing data
  • Fine-mapping of causal variants with targeted sequencing data
  • Application of genotypic data from select ancestry or disease cohorts to enhance ancestry inference and/or disease risk prediction
  • Validation of survey responses through clinical assessment
Who can apply?
  • Institutions can be domestic or foreign, public or private, but must be non-commercial and the research must also be for non-commercial purposes
  • Researchers may be at any stage of their careers. Applications submitted by a group of researchers at the same institution will also be considered
How do I apply?

Your application will include the following information:

  • A research proposal summary, which will be made public if the proposal is accepted (no more than 500 words)
  • A description of the proposed research project (no more than 2,000 words)
  • A description of unique datasets or analytic tools that your research group has developed and will provide for the collaboration. Researchers should include a description of the IRB-approved protocol used to collect the dataset, if applicable (a total of no more than 1,000 words)
  • A list of co-investigators and other personnel
  • An NIH-style biosketch or CV for the principal investigator.

Submit your application using the following form by June 12, 2018Academic Research Collaborations Application Form.

How will applications be evaluated?

Applications will be evaluated by 23andMe scientists based on the following considerations:

  1. Scientific quality and innovativeness of the project proposed
  2. Expertise of and resources available to the investigators
  3. Alignment with 23andMe business objectives
  4. Quality, value and willingness to share complementary datasets
Other Types of Collaborations

The following types of requests are handled on a rolling basis rather than in the review process that occurs twice a year:

  • Requests for access to published datasets. If you are interested in accessing summary statistics from a dataset included in a 23andMe publication, please initiate your request by emailing [email protected], including “Published” and the principal investigator’s last name in the subject line. You will be asked to fill out a short description of your research plan.
  • Requests for replication of GWAS findings. If you have performed a GWAS and would like to test if the association you identified replicate in the 23andMe research cohort, please initiate your request by emailing [email protected], including “Replication” and the principal investigator’s last name in the subject line. You will be asked to fill out an application including details of you study.
Additional Information and Requirements
  • 23andMe data will be shared in aggregate (rather than individual level) and cannot be distributed, shared, or sold to third parties
  • Chosen collaborators will receive non-exclusive access to the data
  • Collaborators must agree that the results of their research will be published in open-access scientific journals (or open-access articles) and according to 23andMe’s publication best practices. Currently, those best practices only allow for the publication of summary statistics for up to 10,000 SNPs; this applies to meta-analyses that include 23andMe data
  • Collaborators may not file for intellectual property rights covering the results of this collaborative research
  • 23andMe may request that collaborators meet with 23andMe (in person or by phone) to discuss the research
  • Please note that this collaboration will not cover the cost of additional data collection or fees for making an article open-access
  • To enable 23andMe’s collaboration and data transfer each collaborator must enter into an Academic Research Agreement with 23andMe, including the above terms and other requirements
About protecting customer privacy

Protecting the privacy of our research participants is of critical importance to 23andMe. Data sharing is structured to provide the same robust privacy protections that 23andMe research participants have as part of their normal participation in 23andMe Research. Only data from 23andMe customers who have accepted 23andMe’s IRB-approved consent document will be incorporated into any analysis. Only de-identified, aggregate data will be shared – no individual level data will be shared. All applicants should take this policy into consideration when designing the research approach.

Any data sharing will take place in accordance with the terms of the 23andMe Academic Research Agreement between a collaborator and 23andMe. Any actions taken by our collaborators that threaten the privacy of 23andMe research participants are grounds for termination of the collaboration and to other remedies described in such agreement.

If you have questions, please contact the Academic Research Collaborations Program Manager Nadia Litterman ([email protected]).