Skip to content

Notebook Gallery

This is the single index for runnable annnet notebooks. The sections follow the work being demonstrated; the Explanations section covers concepts and choices separately.

Base HowTos use the HowTo environment. Scenario and use-case notebooks link their dedicated environments.

Core graph workflows

These notebooks cover construction, graph structure, context, values, and reproducibility.

  • Quickstart


    Build a directed graph, inspect it, and round-trip it through .annnet.

    Open notebook

  • Attributes and views


    Attach metadata to nodes and edges, then inspect dataframe-like views.

    Open notebook

  • Tables and storage


    Build from tables and compare CSV, Excel, native, Parquet, JSON, and NDJSON.

    Open notebook

  • Slices and subgraphs


    Manage context-specific graph state and materialize smaller graphs.

    Open notebook

  • Hyperedges and traversal


    Represent complexes or reactions and traverse local neighborhoods.

    Open notebook

  • Directed hyperedges and stoichiometry


    Work directly with directed hyperedges, incidence coefficients, and reaction-style semantics.

    Open notebook

  • Multilayer


    Work with layers, coupling edges, layer-derived slices, and supra matrices.

    Open notebook

  • Multilayer math


    Inspect supra-adjacency, Laplacians, diffusion, coupling sweeps, and tensor views.

    Open notebook

  • History and reproducibility


    Record mutations, create snapshots, diff graph states, and export history.

    Open notebook

Interoperability and storage

These notebooks show the boundaries where annnet connects to another backend, format, or workflow.

  • Backend accessors


    Inspect optional components and dispatch to installed graph backends.

    Open notebook

  • Flexible edge orientation


    Compare edge-scope and node-scope orientation policies on mixed graphs.

    Open notebook

  • Backend lazy proxies


    Inspect deferred backend conversion and the optional dependency boundary.

    Open notebook

Applied workflows

These larger notebooks combine graph structure with external biological data, optimization, or graph-learning tools.

  • Multi-condition causal signaling


    Fit a signed signaling network to perturbation experiments with CORNETO, then keep the selected network and predicted activities in AnnNet.

    Open notebook · Environment

  • TGF-beta fibrosis response


    Combine signaling, regulation, protein complexes, and metabolism across mechanism and time layers in one AnnNet object.

    Open notebook · Environment

Scenarios

Short ecosystem bridges use small deterministic data and scenario-specific environments.

  • OmniPath table ingestion


    Load an OmniPath-style interaction table as prior knowledge without a network call.

    Open notebook · Environment

  • Cytoscape CX2 export


    Export annotated and hyperedge graphs with explicit CX2 projection modes.

    Open notebook · Environment

  • PyG HeteroData export


    Prepare typed graph data and numeric features for PyTorch Geometric.

    Open notebook · Environment

  • Causal activity bridge


    Store activity scores and causal solution edges from decoupler/CORNETO-style workflows.

    Open notebook · Environment