"New Responsible Data Sharing Technique Will Enable Better Understanding of Disease-Causing Genetic Variants"
Through the use of a novel strategy for securely sharing and analyzing genomic data developed at the UCSC Genomics Institute, scientists may be able to better understand and test for genetic variations that cause cancer and other heritable diseases. To understand the clinical significance of rare genetic variants, large amounts of genomic and clinical data must be analyzed. However, privacy policies limit the sharing of this information between institutions, and no single institution is likely to have all of the resources required for a thorough analysis. UCSC researchers demonstrated how to overcome this problem by "bringing the code to the data" through an approach called federated analysis. This is the first time federated analysis has been used to classify previously unclassified genetic variants. It is essential to find ways to access data that respect privacy while also allowing researchers to conduct research, thus making the federated model the way of the future, according to James Casaletto, the study's lead author. The study concentrated on genetic variants of the breast cancer genes BRCA1 and BRCA2. People who inherit dangerous variants of one of these genes are more likely to develop breast, ovarian, and other cancers. This new study offers a more nuanced understanding of BRCA1 and BRCA2 variants, as well as a proof-of-concept of a novel data sharing and analysis technique for determining the clinical implications of genetic variants. Researchers use the federated analysis approach to bring the code to the data, eliminating the need to export sensitive data. UCSC Genomics Institute software is delivered in a "container" to any collaborating institution worldwide that houses a valuable but protected set of genomic data. The software is then used by the collaborating institution to analyze their data within their secure environment, generating summary data that does not reveal personal information about individual patients. This approach ensures that patient-level data adheres to an institution's strict privacy rules, which prohibit data export, while also allowing researchers to collect a much larger pool of genomic data, leading to better clinical conclusions. This article continues to discuss the use of a federated model to enable a scientific analysis of privacy-sensitive data.