by Abby Olena
The latest project from Google’s high-risk, secretive research arm, Google X, focuses on defining health for human beings. According to the Wall Street Journal, which broke the newsin July, an initial pilot of the so-called “Baseline Study” will collect biological samples from 175 healthy individuals. The samples, which include blood, saliva, urine, and tears, will then be analyzed and the resulting data mined by Google X in order to find genetic and biochemical patterns that correlate with health. ScienceInsider reports that, after the pilot, Google X plans to enroll thousands more participants through collaborations with Duke University and Stanford University.
Reactions to the announcement were mixed. The initial coverage in the WSJ deemed the project a “moonshot,” and other outlets like The Guardian called it “ambitious.” But a piece from Antonio Regalado at MIT Technology Review highlighted the banality of collecting genetic and molecular information and questioned whether the new endeavor qualifies. “As Google itself defines it, a moon shot must combine ‘a huge problem, a radical solution, and the breakthrough technology that might just make that solution possible,’” writes Regalado. “The Baseline study doesn’t fulfill this definition.”
Media coverage also highlighted the ethical issues inherent in any project that collects personal information. The WSJ reported that data will be anonymous, used solely for medical and research purposes, and not shared with participants or insurance companies, but acknowledged that privacy concerns still “lurk in the background.” Stanford radiologist and Baseline Study collaborator Sam Gambhir told the WSJ, “Google will not be allowed free rein to do whatever it wants with this data.” But in Ars Technica, Ron Amadeo pointed out that Google co-founder Larry Page has expressed dismay in the past regarding the restrictions on mining medical data that arise from privacy laws. “With the Baseline Study, it seems that Google intends to build a database of its own that can avoid these limits,” Amadeo wrote.
Based on reports, the health data from volunteers will be collected with the oversight of institutional review boards and will be de-identified before being used. But there are good reasons to doubt that whole genome sequences can ever truly be de-identified. And there are plenty of instances where supposedly anonymous health data were compromised, including 4.5 million records in a single breach this month. Moreover, consumers who invested time and energy uploading their data to Google Health—another voluntary Google database of health information that was discontinued in 2011—might be hesitant the second time around.
Edited by Misha Angrist