sBeacon: Making genomic data-sharing future-ready

An image of pixels all in varying shades of blue. It does not contribute information tot he post and decorative

We’ve made Beacon cloud-native to reduce the cost and processing time of sharing genomic data.

The challenge

Sharing genomic data poses logistical and ethical problems in the research space.

The Beacon protocol has emerged as an international standard to ensure that privacy and data ownership are preserved whilst enabling disease research.

However, the current implementation of Beacon by Elixir is not scalable – the cost associated with querying a schema-based local database increases with cohort size. Consequently, only about 150 Beacon services exist worldwide.

Our response

We collaborated with the Australian Genomics Health Alliance and Genomics England to create a fully cloud-native architecture that enables the sharing of genomic data and the preservation of privacy and data ownership.

Utilising AWS function-compute, Beacon is a fully serverless framework and can be automatically deployed in a user’s cloud account. The framework can run over patient-individual VCF files or large cohort VCF files.

An infographic showing some steps in the query process of sBeacon. The first step is to enter a query, like what is the frequency of genome G at position 16665475 on chromosome 4? SBeacon provides an answer of 0.4 and Step 2 Query, Are there any deletions observed between position 1660000-161000 on chromosome 4? It indicates sBeacon could reply yes or no — sBeacon, cloud-native architecture for faster and more cost-effective genomic data-sharing.

Benefits

Serverless Beacon is 300-fold cheaper and 17-fold faster than the reference Elixir implementation of Beacon.

It does not require data consolidation. This enables distributed multi-institutional collaborations that preserve patient privacy – access to data is file-based and controlled by the data owner.

The framework can segment patient groups on the fly by including only information from relevant individuals, such as filtering by clinical terminology (FHIR).

When not queried, the cloud set-up does not incur cost, making it sustainable for smaller hospitals or research institutes.

Imagine how effective disease research could become if even small hospital or research groups can afford to light up a Beacon and contribute their datasets to the world.

A graph showing a time and cost comparison between Elixir and serverless data exchange. It shows the serverless exchange is faster and cheaper. — Serverless Beacon is cheaper and faster than the current implementation of Beacon by Elixir.

You can follow live updates here.