The task
IMTM set out to build a comprehensive Czech genome database through the ENIGMA project, processing 4,000+ sequenced genomes from Czech donors to determine population-specific characteristics of genetic variants.
- Data volume: traditional approaches could not manage the dataset.
- Storage and accessibility: existing systems lacked efficient storage, querying and retrieval for genomic data.
- Resource constraints: in-house development would need substantial computing resources and staff.
What we built
Floxgen Aggregation automates genomic data processing, turning raw variants into actionable insight with Apache Hail and ElasticSearch. One-click recalculation updates everything whenever new samples are added or specific populations need to be analysed.
The workflow starts with uploading VCF files, followed by ingestion into a big-data database and pre-processing. Plugins annotate variants for clinical context; aggregation and statistics reveal patterns and associations. Scientists access the data through ElasticSearch, a relational database or a GraphQL API, and explore it in a web portal.
The result
- Development time down by up to 90% versus traditional methods
- Significantly lower cost of building the database
- Lower technological and staffing barriers to population databases
- The Czech Genome Portal, powered by Floxgen, gives researchers and clinicians insight for personalized healthcare