How we've made a global search engine for genetic data
summary
A federated search called Beacon Network maps queries across over 100 beacons from 40 institutions in 18 countries, touching an estimated 2 million samples and over 2 billion variants. The system implements a GA4GH standard, runs as nine subsystems (18 modules) on Docker/Kubernetes with a four-stage execution pipeline (converter, requester, fetcher, parser) and many adapters (15 converters, 11 requesters, 18 parsers). Adoption and usefulness are validated by roughly 1–1.5 million searches.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:00The Beacon Network
- 7:46Why is it needed
- 8:50What it means
- 9:42The Beacon Project
- 11:48Beacon Network
- 17:55System structure
- 25:47Statistics
- 29:26Properties
- 32:20Variants
- 33:24Diseases
related talks
26mBuilding a scalable genomic matching algorithm
05mDuckDB powering cancer genomics research
28mBuilding a Cutting Edge Search Engine in Go
19mStandard FHIR clinical genomic operations for genetic data
29mAnalysing and sharing genetic data with Python
46mBiomedical Data Translator - Connecting the Dots
38mBuild Gen AI applications with Google Databases and Cloud Runtimes
31mReliability and Resilience in the Cloud