DIRAC
DIRAC is a spatial omics tool whose code is published by boxiangliulab on GitHub, first public in 2026, described in Nature Methods (2026). Graph neural network that integrates spatial multi-omic data into a domain-invariant embedding and automates cell-type annotation by transferring labels across studies.
| Category | Spatial omics |
|---|---|
| Developer (GitHub) | boxiangliulab |
| Task | Vertical and horizontal integration; automated cell-type annotation via label transfer; spatial domain identification |
| Interface | Python package (PyTorch, Scanpy); pip install sodirac |
| Website | dirac-tutorial.readthedocs.io/en/latest |
| GitHub | boxiangliulab/DIRAC |
| Released | 2026-11-05 (source) |
| Paper | https://doi.org/10.1038/s41592-026-03282-9 · Nature Methods (2026) |
| Code license | AGPL-3.0 |
| License note | dual licence: AGPL-3.0 for non-commercial/academic use; commercial licence available from the authors |
| What this allows | Non-commercial Non-commercial use such as academic research is OK; commercial use needs a separate license from the owner. What do these mean? |
| License source | github.com/boxiangliulab/DIRAC/blob/main/LICENSE · checked 2026-11-08 |
| Pricing | Open source (non-commercial) |
| Tags | #spatial-multiomics#integration#graph-neural-network#cell-type-annotation |
| Last checked | · links verified |
Can I use DIRAC commercially?
No. The recorded terms do not allow commercial use without a separate license from the owner.
- Code: AGPL-3.0
- Source: github.com/boxiangliulab/DIRAC/blob/main/LICENSE · license checked 2026-11-08
What the license types mean · not legal advice.
Specs
| Platforms supported | Sequencing-based and imaging-based spatial multi-omics, including Stereo-CITE-seq |
|---|---|
| Resolution | — |
| Tasks | Vertical and horizontal integration; automated cell-type annotation via label transfer; spatial domain identification |
| GitHub stars | 5 |
| Interface | Python package (PyTorch, Scanpy); pip install sodirac |
Compare with…
Pick 1–3 other Spatial omics tools to compare side by side with DIRAC.
Related
Spatial single-cell analysis toolkit in the scverse ecosystem.
Spatial omicsOpen sourcePermissive (commercial OK)#python
Robust cell type decomposition for spatial transcriptomics.
Spatial omicsOpen sourceCopyleft (commercial OK, share-alike)#deconvolution
Bayesian mapping of fine-grained cell types in spatial transcriptomics.
Spatial omicsOpen sourcePermissive (commercial OK)#deconvolution