SAN FRANCISCO (GenomeWeb) – One scientist's quest to make use of direct-to-consumer genomic tools to piece together health implications of his family's genomes has led to renewed questions about data sharing, consent, and what to do with the genomic data of the deceased.
Manuel Corpas, founder and CEO of Cambridge Precision Medicine, began analyzing and making public his family's genomic data several years ago when he first ordered 23andMe kits for himself, his sister, and parents and published the analysis in the journal F1000 Research in 2012. He followed that up with the sequencing and analysis of the family's exomes using only publicly available tools in an attempt to compare various analysis platforms and also to see how an individual might access and use his or her own genomic data.
Following his aunt's death from melanoma, he wanted to take the project a step further and see whether he could identify anything in the aunt's genome that would have increased her risk for cancer, and thus be informative to relatives. But in doing that, the project began to raise the general question of what should be done with genomic data of the deceased.
One issue that became clear when he and his family sequenced his deceased aunt's genome is that as sequencing prices drop and population sequencing projects pick up, there will be vast amount of data generated. But, he said, "eventually, people will pass away, and we haven't had any serious discussion about how to manage this type of information, which could be extremely valuable," both for research and potentially for individuals' family members, he said.
Enabling people to donate their genomic data raises some interesting ethical and logistical hurdles. Mark Rothstein, founding director of the Institute for Bioethics, Health Policy and Law at the University of Louisville School of Medicine, is collaborating with Bartha Knoppers, director of the Center of Genomics and Policy at McGill University, to study this very issue.
For genomic datasets that have already been generated, "one reason there is little sharing is because these samples have value and the institutions that are in possession of them don't want to give away a valuable resource," he said.
Irrespective of the monetary value, however, in order to make such data truly valuable, a lot of investment would be needed to curate them and then, to make the samples really valuable, they would need to be linked with medical records. "That raises all sorts of additional privacy issues and issues around informed consent," he said.
The Personal Genomes Project in the UK is grappling with the issue of genomic data donation. The project is working to put in place a mechanism by which not only participants in the project, but others who have had their genomes sequenced, can deposit the data in a database that is freely accessible to researchers.
"There are a large number of genomes being generated under a consent form that doesn't allow the data to be openly available," Stephan Beck, professor of medical genomics at the University College London Cancer Institute, said. "It would be nice to find a way for data that's already out there and for which the individuals want it to be available, to make it available," he said. Currently, however, there is not a mechanism to enable that. "We have mechanisms in place to donate organs, blood, and tissue, but not genomic data," he said.
Beck and others, as part of the UK PGP, are looking to put in place such a mechanism, which they've called an "open consent" agreement. "It's based on the fundamental principal that the individual whose genome has been sequenced for whatever reason is the final owner of their data," he said.
So far, five individuals have contributed genomic data to the Personal Genomes Project database. Beck said that the team has not yet made a large-scale effort to recruit genome donations because it is still working through the kinks, trying to anticipate potential problems. In addition, he said, the team is developing automation tools.
"Before we open the floodgates we want to make sure we have the pipeline in place," Beck said.
The first person to donate his genome to the UK PGP was Colin Smith, a professor of functional genomics at the University of Brighton in the UK, who previously had his genome sequenced as part of Illumina's Understand Your Genome symposium in 2013. In a blog post on the PGP website, Smith discussed his motivation for donating his data and said that since making his data public, he's followed up with his doctor about several serious health risks that were uncovered in his genome.
Rothstein said that aside from the open consent model of the UK PGP, other types of consent models have been proposed. For instance, in the US, the National Institutes of Health's All of Us project will seek to gather genomic and health data from 1 million individuals. In addition, he said there are models being proposed and developed by organizations like the Genetic Alliance and Patients Like Me that are more patient-driven and oriented, rather than research oriented. Rothstein is currently co-leading a three-year study funded by the NIH that is evaluating the intersection of two growing trends: research being conducted by "citizen scientists" and unregulated research taking place on data collected from mobile devices.
These days, he said, anyone can conduct genomics research on themself. You can have yourself sequenced by DTC labs, join an online network of people with the same disease or same genetic variant, and that's not regulated, Rothstein said, which raises concern about informed consent. In addition, he said, there are a growing number of health-related cell phone apps that enable blood testing and electrocardiograms, for example.
"These are regulated to a degree, but there's a lot of misinformation associated with these apps," he said. "When you download them, you consent to give away everything to the app developers, and they are free to sell your medical information."
When thinking about the future of genomic sequencing and making genomic data public and accessible for research, Rothstein said that such consent and privacy issues will need to be addressed. Genome data donation is an idea being explored by a number of groups, he added, and there have been proposals for "genomic trusts, where a trusted third party is responsible for collecting the data and parceling it out to registered researchers," he said.
Corpas said that the UK PGP model for genome donation was a good starting point, but added that there are other factors to consider as well. For instance, he said, all existing frameworks tend to fall short in considering differing preferences within individuals of the same family.
"The consent frameworks work well when you're talking about your own genome or the genome of someone you're the guardian for, but when talking about how the different consents from different family members can be reconciled, that's a challenge," he said.
In Corpas' family's case, he had the consent of his sister and parents to sequence and make available their genomic data, and he also had the consent of his uncle to sequence his aunt's genome and make that information public. But had his mother, for instance, not wanted to know her genomic data, it's unclear what Corpas' responsibility would be if he identified a hereditary cancer risk variant in his aunt's genome, since that would have implications for his mother, as well as himself and his sister.
"I want this debate to take place," he said.