Tutorials

These notebooks show annnet in a larger analysis or at the boundary with another tool. For the graph model and everyday API usage, start with the Guide.

Case studies

Notebook	Analysis	Environment
Multi-condition causal signaling	Fit a small signaling network with CORNETO; keep nine synthetic perturbation experiments, selected interactions, and predictions in AnnNet.	YAML
TGF-beta fibrosis response	Combine signaling, regulation, complexes, and metabolism across time; join measurements, trace paths, export to Cytoscape, and forecast responses.	YAML

Integrations

Notebook	What it demonstrates	Environment
OmniPath table ingestion	Import a local OmniPath-style table through the knowledge-base client and add analysis slices.	YAML
Cytoscape CX2 export	Choose a hyperedge projection and round-trip the embedded AnnNet manifest.	YAML
PyG HeteroData export	Convert a heterogeneous graph to tensors for PyTorch Geometric.	YAML
Causal activity bridge	Store a CORNETO CARNIVAL solution, selected interactions, and activities alongside the prior graph.	YAML

Reproduce a tutorial

Use the notebook's linked environment rather than installing every integration into the guide environment. For example:

conda env create -f docs/tutorials/environments/cytoscape-cx2-export.yml

Activate the environment named in that YAML, then open the notebook in Jupyter. Case studies may require downloaded data, cached input files, and a solver; read their setup cells before running them. The website renders saved outputs for these six tutorials. Its documentation build executes only the five Guide notebooks, so a successful site build does not validate external services or rerun the case studies.

The fibrosis case study currently needs prepared inputs that are not bundled with this repository: measured.parquet, responsive.parquet, the cached signaling.cx2 backbone, dorothea.tsv, omnipath_complexes.tsv, Human-GEM.xml, and proteinatlas.tsv. Its saved outputs can be read here, but the environment alone is not sufficient to reproduce it.