Confidential computing network for cross-database DNA relatives matching

Le résumé fourni par la source

Genomic data is continuously collected by different entities and thus siloed by default, hampering collaborative research. Meanwhile, relatedness inference, a basic component of genomic analysis involving only genomic data itself, presents a massive usage in consumer genomics, where millions of individuals willingly share their data with various services in hopes of discovering familial connections, resulting in extensive genomic data sharing. This work presents a pioneering approach for secure and privacy-preserving genomic data sharing leveraging both GA4GH standards as a fundamental basis for interoperability and a novel privacy-enhancing Confidential Computing technology. Unlike conventional computing environments, Confidential Computing protects data while it's in use by performing computations in hardware-based, isolated, and verifiable Trusted Execution Environments (TEEs). In practice, TEE may allow for the verification of data access policies and minimisation of data exposure efficiently to zero, even when data is shared with remote, uncontrolled information systems. This ensures trusted, consented processing, yielding insights such as the detected degree of kinship. While initially addressing the consumer genomics use case, the presented approach has the potential to be extended for research purposes.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.