CHICAGO – A coalition of researchers and healthcare providers in Canada and beyond are building what they believe to be the first global, federated network for sharing genomics and clinical data to support advances in research and patient care for autism.
Called the Autism Sharing Initiative, the program will follow standards developed by the Global Alliance for Genomics and Health, or GA4GH, to facilitate cross-institutional collaboration on research into genetic factors related to autism. Users will be able to search and analyze de-identified datasets hosted in disparate locations as well as to access individual patient data for treatment, with proper patient consent.
"The goal will be to create the world's largest federated network of genomics and clinical data in autism," said Marc Fiume, CEO of DNAstack, a Toronto-based bioinformatics company that leads the project.
"It's going to take big data to understand autism," added Dean Hartley, senior director of discovery and translational science for Autism Speaks. "One of the problems is that a lot of autism data sits in silos at a lot of academic institutions."
The autism advocacy group is trying to get the most complete picture of patients as possible in front of any researcher who wants it. According to Hartley, federation is the key mechanism to break down these silos, and cloud computing and standards facilitate this.
This federated model distinguishes the Autism Sharing Initiative from other large-scale research projects such as the Simons Foundation's SPARK program, where all the sequencing was performed by two laboratories.
Hartley explained that a federated system can help researchers find cohorts with rare subtypes of autism or any other condition with a genetic basis. "Cancer is way ahead of us in terms of how [to] get other omics into the profile with the genomics and the phenotypic profiling," he said.
One partner in the Autism Sharing Initiative, Molecular You, is going to bring proteomics and metabolomics into the mix. That Vancouver, British Columbia-based company has developed an artificial intelligence-driven system that provides multidimensional analysis of patient data, including data from microbiomic, metabolomic, proteomic, and environmental exposure testing, and clinical sources to help individuals manage their health.
The Autism Sharing Initiative is one of 24 "driver" projects of GA4GH, meant to test the group's standards in real-world settings. Other such projects include All of Us, Genomics England, the Clinical Genome Resource, and the Human Cell Atlas, though Fiume said that the autism project is the largest and most robust implementation of GA4GH standards to date.
"For us, it feels like the holy grail of the GA4GH is to enable distributed data sharing, discovery access, and machine learning, where data can live in different environments and different geographies, but still kind of feel to the researcher like they're operating on one," Fiume said.
Fiume said that the Autism Sharing Initiative will be a major validation of GA4GH standards and principles. "This is kind of enabling an internet of genomics data," he said.
The Autism Sharing Initiative has received C$11.2 million ($9.2 million) in support from its partners, including C$4.3 million from the Digital Technology Supercluster, a federally funded consortium across Canada focused on advancing the country's efforts to develop and apply digital technologies to challenges in natural resources, healthcare, and industry.
The initial funding period is 18 months, though Fiume said that he expects the program to continue long past then. "I would say we're not going to be done for a few years," he said.
The Digital Technology Supercluster convenes collaborative R&D efforts so participants can conceive and develop their concepts, then the organization funds as much as 50 percent of the work. In healthcare, the consortium specifically concentrates on the application of data to enable precision health.
Other partners in the consortium include Autism Speaks and Autism Speaks Canada, Excelar Technologies, the McGill University Centre of Genomics and Policy, Molecular You, the Pacific Autism Family Centre Foundation, Roche Canada, the Hospital for Sick Children in Toronto (SickKids), and the University of British Columbia. The initiative also is receiving support from the Holland-Bloorview Kids Rehabilitation Hospital, the Ontario Brain Institute, and the BC Children's Hospital Research Institute's Autism Spectrum Interdisciplinary Research program.
"For us, this represents the opportunity to demonstrate a precision health pipeline," Fiume said. "We've brought together a wonderful consortium from patients and families, academic research, patient advocacy, clinical research, software companies, and pharma."
Many of the participants have worked together in the past.
Bill Tam, cofounder of the Vancouver-based Digital Technology Supercluster, said that there is a "solid foundation of trust" among the partners, though autism represents a new set of needs in that the "patient journey" is so central.
Hartley said that DNAstack introduced Autism Speaks to GA4GH, and those standards have helped the advocacy group to understand some of the technical considerations behind data sharing, including user authorization and privacy. The supercluster is also assisting with regulatory, ethical, and policy issues, including patient consent, he said.
DNAstack has worked with Autism Speaks since 2018, initially to restructure the advocacy group's massive Mssng research portal, with the goal of simplifying access to genomic data on autism and ultimately accelerating research through open science.
Mssng, pronounced "missing," is an ongoing effort between Autism Speaks, SickKids, and Google and contains an open repository of whole-genome sequence, phenotype, and clinical information from more than 10,000 individuals and families with autism. The database and associated analytics tools are offered openly to researchers, clinicians, genetic counselors, and families affected by the condition to help fill in knowledge gaps.
Last winter, the Digital Technology Supercluster funded a DNAstack-led partnership to develop COVID Cloud, a bioinformatics and health data platform for COVID-19 research that also follows GA4GH standards. COVID Cloud itself builds upon Beacon for SARS-CoV-2, which the firm released early in the pandemic to enable the scientific and medical communities to share and discover knowledge about the genetics of the virus in real time.
The Autism Sharing Initiative is "very much an evolution of Beacon," an application programming interface for data sharing from GA4GH, Fiume said. He described DNAstack as the technical backbone of the Autism Sharing Initiative.
"We've been incrementally building technology to enable collaborative biomedical research," Fiume said. "We, at a very high level, are trying to build the technical bridges that make data sharing and discovery through shared data much more efficient."
Tam noted that the Autism Sharing Initiative is distinct from those earlier efforts because it involves patient-specific data as well as anonymized records. "I think the patient journey on this is what will make the big difference here," he said, largely because autism is a lifelong condition.
Indeed, Autism Speaks is setting up a community portal as well as a research portal for the initiative to get families affected by autism involved and connected to each other.
"We hope that by building community portals … families can give information back to researchers, or researchers can interface with physicians and physicians can interface with the families," Hartley said.
Fiume said that DNAstack will be launching four products for the Autism Sharing Initiative this year, starting in the fall: one for researchers to publish their genomics and clinical data, a search engine for shared data, another app for identity verification, and a fourth that will contain a suite of analytics tools.
Hartley added that this model could be replicated in other genetic diseases, though he did not specify any. "I do have a strong focus and hope that it will help to understand autism in the future, but I also do hope that [it will serve as] a greater good for others to see that these tools or platforms [can] be used in a similar way that helps them," Hartley said.