Typical analysis patterns
Most workflows in annnet fall into a small number of patterns.
Multilayer analysis
Use this when layers are part of the scientific model, for example modalities, states, or timepoints.
- Typical flow: define aspects, add layer tuples, add intra-layer and inter-layer edges, then analyze through annnet's multilayer operators or an external backend.
- Read first: Multilayer and multi-aspect graphs.
- Useful API pages: Layers, Graph.
- Notebooks: SBUC, HEK293 multilayer network.
Hyperedges and stoichiometry
Use this when binary edges are too limiting and you need directed hyperedges, weighted memberships, or SBML-style reaction structure.
- Typical flow: import SBML, inspect hyperedges, preserve stoichiometric weights, then analyze or export.
- Read first: Incidence representation.
- Useful API pages: Graph, SBML.
- Notebook: AnnNet Showcase.
Slices and scenario management
Use this when you need multiple contexts in one container, such as train/eval splits, perturbation scenarios, or condition-specific subgraphs.
- Typical flow: define slices, switch the active slice, override per-slice weights or annotations, then compare results.
- Read first: Slices and views.
- Useful API pages: Slices, Views.
- Also useful: Tracking changes.
Interoperability-first analysis
Use this when annnet is the main container and you want to move into other tools without losing structure.
- Typical flow: convert with
an.to_nxoran.to_igraph, compute or edit, restore with the manifest, then save as.annnet. - Read first: Interoperability.
- Useful API pages: NetworkX adapter, igraph adapter, Native .annnet format.
- For persistence and exchange formats: Storage and IO.
Notebook picks
- Start with AnnNet Showcase if you want the broadest introduction.
- Use Multilayer Systems Biology if multilayer biology is your main use case.
- Use SBUC without Polars if you want the same example in a lighter environment.
- Use AnnNet Demo if you want a shorter hands-on notebook.
For a broader tour of examples, open the Notebook gallery.