tiledbvcf

You can store and query VCF files for cohort studies efficiently by ingesting indexed single-sample VCF/BCF files into a TileDB sparse-array dataset, enabling rapid region- and sample-scoped reads without costly re-merging. This allows high-throughput querying across many samples.

Adding new genomic samples to an existing variant dataset is handled through incremental ingestion. You ingest single-sample indexed VCF/BCF files and add new samples later without expensive rebuilding, supporting iterative sample expansion for population genomics workflows.

To extract specific genomic regions and attributes from a VCF cohort, query the TileDB sparse-array dataset by genomic windows and sample subsets while selecting specific attributes like genotypes and positions, then export selected regions back to VCF or tabular outputs.

Yes, you need correctly indexed VCF/BCF files (.csi or .tbi) to ingest genomic variant data into a sparse array. The ingestion process requires single-sample VCF inputs and supports parallel ingestion and compressed storage for high-throughput querying.

Yes, you can export queried variant subsets back to VCF format or tabular outputs. This interoperable export capability supports sharing, QC, and downstream analysis workflows like association testing after selecting specific regions and samples.

Querying large VCF datasets requires a sparse-array storage model to enable high-performance region- and sample-specific queries. This approach avoids slow filesystem-style access and costly re-merging, applying compressed storage and flexible attribute selection for cohort-scale analysis.