Extract knowledge assertions from tabular data into NCATS Translator-compliant KGX NDJSON — declaratively, with entity resolution and quality control built in.
pip install tablassert
tablassert build-knowledge-graph config.yamlFull Documentation — installation guides, tutorials, configuration reference, and API docs.
pip install tablassertAll dependencies (ML, web, Excel support) are included in the base install. An optional extra is available for CPU compatibility:
pip install "tablassert[rtcompat]" # Polars build for CPUs without required instructionsDocker
docker pull ghcr.io/skyeav/tablassert:latest
docker run --rm \
-v /path/to/config:/data \
-v /path/to/datassert:/datassert \
ghcr.io/skyeav/tablassert:latest \
build-knowledge-graph /data/graph-config.yaml# Build a knowledge graph from a YAML configuration
$ tablassert build-knowledge-graph graph-config.yaml
⠋ Loading table configurations...
⠋ Resolving entities across 16 DuckDB shards...
⠋ Compiling subgraphs...
⠋ Deduplicating nodes and edges...
✓ Done — wrote nodes.ndjson and edges.ndjson to .storassert/Define your entities and relationships in YAML, point tablassert at your data, and get NCATS Translator-compliant KGX NDJSON out the other side — no code required.
- Declarative Configuration — YAML-based, no code required
- Entity Resolution — Maps text to biological entities (genes, diseases, chemicals)
- Quality Control — Three-stage validation (exact → fuzzy → BERT embeddings)
- KGX Compliance — NCATS Translator-compatible NDJSON output
- Performance — Lazy evaluation pipelines with Polars and DuckDB-accelerated entity resolution
See CONTRIBUTING.md for development setup, code style, and pull request guidelines.
Skye Lane Goetz — Institute for Systems Biology, CalPoly SLO
Gwênlyn Glusman — Institute for Systems Biology
Jared C. Roach — Institute for Systems Biology