Chunk is a computational framework designed to leverage phenotype information from bulk transcriptomic data to uncover robust associations between cell-cell interactions (CCIs) and clinical or biological phenotypes in single-cell or spatial transcriptomic data.
🧠 Core Hypothesis: Intercellular communication mediated by ligand–receptor interactions (LRIs) drives phenotypic heterogeneity across patients.
Guided by diverse phenotypic data types (binary, linear, ordinal, survival), Chunk identifies phenotype-associated LRIs from large-scal 6D8B e bulk cohorts and maps them to the single-cell or spatial level to uncover CCI events associated with disease-related phenotypic variation.
Chunk is implemented in Python 3 and can be installed via:
pip install chunk-pyExplore how to apply Chunk to various datasets and phenotype types:
🔍 Note: Spatial transcriptomic analysis in Chunk is fundamentally similar to single-cell analysis. For example:
- To conduct binary phenotype + spatial analysis, combine:
- The first half of the binary + single-cell tutorial
- The second half of the ordinal + spatial tutorial.
| Phenotype Type | Dataset Type | Notebook Link |
|---|---|---|
| Binary | Single-cell | 🔗 View Tutorial |
| Linear | Single-cell | 🔗 View Tutorial |
| Ordinal | Spatial | 🔗 View Tutorial |
| Survival | Single-cell | 🔗 View Tutorial |
You can download the example dataset for tutorials here: https://drive.google.com/drive/folders/17RgFhzNYNzFHYUq1Oo0bjhOZDNkfUtff?usp=sharing
If you use Chunk in your research, please consider citing our paper (coming soon).