Continued advances in precision medicine rely on the widespread sharing of data that relate human genetic variation to disease. However, data sharing is severely limited by legal, regulatory, and ethical restrictions that safeguard patient privacy. Federated analysis addresses this problem by transferring the code to the data-providing the technical and legal capability to analyze the data within their secure home environment rather than transferring the data to another institution for analysis. This allows researchers to gain new insights from data that cannot be moved, while respecting patient privacy and the data stewards' legal obligations. Because federated analysis is a technical solution to the legal challenges inherent in data sharing, the technology and policy implications must be evaluated together. Here, we summarize the technical approaches to federated analysis and provide a legal analysis of their policy implications.

The UCSC Cancer Genomics Browser (https://genome-cancer.ucsc.edu) offers interactive visualization and exploration of TCGA genomic, phenotypic, and clinical data, as produced by the Cancer Genome Atlas Research Network. Researchers can explore the impact of genomic alterations on phenotypes by visualizing gene and protein expression, copy number, DNA methylation, somatic mutation and pathway inference data alongside clinical features, Pan-Cancer subtype classifications and genomic biomarkers. Integrated Kaplan-Meier survival analysis helps investigators to assess survival stratification by any of the information.

Genetic tests of the cancer predisposition genes BRCA1 and BRCA2 inform significant clinical decisions for both physicians and patients. Most uncovered variants are benign, and determining which few are pathogenic (disease-causing) is sometimes challenging and can potentially be inconsistent among laboratories. The ClinVar database makes de-identified clinical variant classifications from multiple laboratories publicly available for comparison and review, per recommendations of the American Medical Association (AMA), the American College of Medical Genetics (ACMG), the National Society for Genetic Counselors (NSGC), and other organizations. Classifications of more than 2000 BRCA1/2 variants in ClinVar representing approximately 22,000 patients were dichotomized as clinically actionable or not actionable and compared across up to seven laboratories. The properties of these variants and classification differences were investigated in detail. Per-variant concordance was 98.5% (CI 97.9%-99.0%). All discordant variants were rare; thus, per patient concordance was estimated to be higher: 99.7%. ClinVar facilitated resolution of many of the discordant variants, and concordance increased to 99.0% per variant and 99.8% per patient when reclassified (but not yet resubmitted) variants and submission errors were addressed. Most of the remaining discordances appeared to involve either legitimate differences in expert judgment regarding particular scientific evidence, or were classifications that predated availability of important scientific evidence. Significant classification disagreements among the professional clinical laboratories represented in ClinVar are infrequent yet important. The unrestricted sharing of clinical genetic data allows detailed interlaboratory quality control and peer review, as exemplified by this study.

Background

Genetic tests of the cancer predisposition genes BRCA1 and BRCA2 inform significant clinical decisions for both physicians and patients. Most uncovered variants are benign, and determining which few are pathogenic (disease-causing) is sometimes challenging and can potentially be inconsistent among laboratories. The ClinVar database makes de-identified clinical variant classifications from multiple laboratories publicly available for comparison and review, per recommendations of the American Medical Association (AMA), the American College of Medical Genetics (ACMG), the National Society for Genetic Counselors (NSGC), and other organizations.

Methods

Classifications of more than 2000 BRCA1/2 variants in ClinVar representing approximately 22,000 patients were dichotomized as clinically actionable or not actionable and compared across up to seven laboratories. The properties of these variants and classification differences were investigated in detail.

Results

Per-variant concordance was 98.5% (CI 97.9%-99.0%). All discordant variants were rare; thus, per patient concordance was estimated to be higher: 99.7%. ClinVar facilitated resolution of many of the discordant variants, and concordance increased to 99.0% per variant and 99.8% per patient when reclassified (but not yet resubmitted) variants and submission errors were addressed. Most of the remaining discordances appeared to involve either legitimate differences in expert judgment regarding particular scientific evidence, or were classifications that predated availability of important scientific evidence.

Conclusions

Significant classification disagreements among the professional clinical laboratories represented in ClinVar are infrequent yet important. The unrestricted sharing of clinical genetic data allows detailed interlaboratory quality control and peer review, as exemplified by this study.

The Cancer Genomics Hub (CGHub) is the online repository of the sequencing programs of the National Cancer Institute (NCI), including The Cancer Genomics Atlas (TCGA), the Cancer Cell Line Encyclopedia (CCLE) and the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) projects, with data from 25 different types of cancer. The CGHub currently contains >1.4 PB of data, has grown at an average rate of 50 TB a month and serves >100 TB per week. The architecture of CGHub is designed to support bulk searching and downloading through a Web-accessible application programming interface, enforce patient genome confidentiality in data storage and transmission and optimize for efficiency in access and transfer. In this article, we describe the design of these three components, present performance results for our transfer protocol, GeneTorrent, and finally report on the growth of the system in terms of data stored and transferred, including estimated limits on the current architecture. Our experienced-based estimates suggest that centralizing storage and computational resources is more efficient than wide distribution across many satellite labs. Database URL: https://cghub.ucsc.edu.

The University of California Santa Cruz (UCSC) Genome Browser (http://genome.ucsc.edu) offers online public access to a growing database of genomic sequence and annotations for a large collection of organisms, primarily vertebrates, with an emphasis on the human and mouse genomes. The Browser's web-based tools provide an integrated environment for visualizing, comparing, analysing and sharing both publicly available and user-generated genomic data sets. As of September 2013, the database contained genomic sequence and a basic set of annotation 'tracks' for ∼90 organisms. Significant new annotations include a 60-species multiple alignment conservation track on the mouse, updated UCSC Genes tracks for human and mouse, and several new sets of variation and ENCODE data. New software tools include a Variant Annotation Integrator that returns predicted functional effects of a set of variants uploaded as a custom track, an extension to UCSC Genes that displays haplotype alleles for protein-coding genes and an expansion of data hubs that includes the capability to display remotely hosted user-provided assembly sequence in addition to annotation data. To improve European access, we have added a Genome Browser mirror (http://genome-euro.ucsc.edu) hosted at Bielefeld University in Germany.

The Cancer Genomics Hub (CGHub) is the online repository of the sequencing programs of the National Cancer Institute (NCI), including The Cancer Genomics Atlas (TCGA), the Cancer Cell Line Encyclopedia (CCLE) and the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) projects, with data from 25 different types of cancer. The CGHub currently contains >1.4 PB of data, has grown at an average rate of 50 TB a month and serves >100 TB per week. The architecture of CGHub is designed to support bulk searching and downloading through a Web-accessible application programming interface, enforce patient genome confidentiality in data storage and transmission and optimize for efficiency in access and transfer. In this article, we describe the design of these three components, present performance results for our transfer protocol, GeneTorrent, and finally report on the growth of the system in terms of data stored and transferred, including estimated limits on the current architecture. Our experienced-based estimates suggest that centralizing storage and computational resources is more efficient than wide distribution across many satellite labs. Database URL: https://cghub.ucsc.edu.

The University of California Santa Cruz (UCSC) Genome Browser (http://genome.ucsc.edu) offers online public access to a growing database of genomic sequence and annotations for a large collection of organisms, primarily vertebrates, with an emphasis on the human and mouse genomes. The Browser's web-based tools provide an integrated environment for visualizing, comparing, analysing and sharing both publicly available and user-generated genomic data sets. As of September 2013, the database contained genomic sequence and a basic set of annotation 'tracks' for ∼90 organisms. Significant new annotations include a 60-species multiple alignment conservation track on the mouse, updated UCSC Genes tracks for human and mouse, and several new sets of variation and ENCODE data. New software tools include a Variant Annotation Integrator that returns predicted functional effects of a set of variants uploaded as a custom track, an extension to UCSC Genes that displays haplotype alleles for protein-coding genes and an expansion of data hubs that includes the capability to display remotely hosted user-provided assembly sequence in addition to annotation data. To improve European access, we have added a Genome Browser mirror (http://genome-euro.ucsc.edu) hosted at Bielefeld University in Germany.

Testing for variation in BRCA1 and BRCA2 (commonly referred to as BRCA1/2), has emerged as a standard clinical practice and is helping countless women better understand and manage their heritable risk of breast and ovarian cancer. Yet the increased rate of BRCA1/2 testing has led to an increasing number of Variants of Uncertain Significance (VUS), and the rate of VUS discovery currently outpaces the rate of clinical variant interpretation. Computational prediction is a key component of the variant interpretation pipeline. In the CAGI5 ENIGMA Challenge, six prediction teams submitted predictions on 326 newly-interpreted variants from the ENIGMA Consortium. By evaluating these predictions against the new interpretations, we have gained a number of insights on the state of the art of variant prediction and specific steps to further advance this state of the art.

The UCSC Genome Browser (https://genome.ucsc.edu) is a web-based genomic visualization and analysis tool that serves data to over 7,000 distinct users per day worldwide. It provides annotation data on thousands of genome assemblies, ranging from human to SARS-CoV2. This year, we have introduced new data from the Human Pangenome Reference Consortium and on viral genomes including SARS-CoV2. We have added 1,200 new genomes to our GenArk genome system, increasing the overall diversity of our genomic representation. We have added support for nine new user-contributed track hubs to our public hub system. Additionally, we have released 29 new tracks on the human genome and 11 new tracks on the mouse genome. Collectively, these new features expand both the breadth and depth of the genomic knowledge that we share publicly with users worldwide.