Biotope

Biotope turns reviewed descriptions of local data into typed Python graph projects. You define the research purpose, source readers, identities and transformations; Biotope checks the definitions and exports BioCypher files with provenance.

Start here

• Install Biotope from PyPI, including its published croissant-baker dependency.
• Run the tutorial to build a small graph from synthetic CSV data.
• Author a graph from your own curated source descriptions.
• Set up a coding agent to use the same workflow with repository skills.

Workflow

Local data → Croissant metadata → Review → Source inventory
                                                ↓
                  Python schemas, loaders, topology and mappings
                                                ↓
                              Check → Build → Review graph output

biotope add reads local data and describes it with croissant-baker. Review those metadata before generating source classes: a successful scan does not establish that every field or file was described completely.

biotope graph scaffold creates a workspace, and biotope source generate gives every described input a source package with a schema to edit. Project Python code then loads selected data, transforms typed records and constructs the graph. biotope graph check checks declarations and types; biotope graph build executes the pipeline, validates its objects and writes the export.

A build records source references, exclusions and the concept and property descriptions that interpret the graph. Use these to assess whether the graph supports its intended questions. Structural checks alone cannot establish scientific validity: review the graph against the sources themselves.

Reference

• Commands
• Architecture
• Technical notes for typed graphs
• Shared annotation policies
• Migrating to 0.10

Biotope 0.10 supports Python 3.10–3.12. APIs may change while the project is under active development. Source code and issue reporting are on GitHub.