Glossary

Keep this page handy when reading the tutorials or API docs. It collects the recurring concepts in Coord2Region and points to the sections that explain them in depth.

Atlas Fetcher

The utility class (coord2region.fetching.AtlasFetcher) responsible for downloading, caching, and managing atlas files from remote repositories. It handles the logic for storing templates in your home directory (default: ~/nilearn_data).

Atlas Mapper

The core engine (coord2region.coord2region.AtlasMapper) that performs the mathematical translation between 3D coordinates and anatomical labels. See documentation/atlases for the list of supported parcellations.

Builder

The visual configuration editor hosted on GitHub Pages. It allows you to design pipelines interactively and export the YAML configuration without writing code. See documentation/web_interface.

Config File

A YAML file (typically coord2region.yaml) that defines a reproducible run. It captures inputs, enabled atlases, API keys, and output settings in a single document. See documentation/cli_tools.

MNI152

The standard brain template (Montreal Neurological Institute) used by Coord2Region. All input coordinates (x, y, z) are assumed to be in MNI space unless otherwise specified.

NiMARE

The underlying library used to interface with neuroimaging meta-analysis databases like Neurosynth and NeuroQuery.

Pipeline

The sequential workflow that chains distinct modules together: Coordinate $rightarrow$ Atlas Lookup $rightarrow$ Literature Search $rightarrow$ AI Summary. See Concepts & Workflow for the logic breakdown.

Provider

Any external service or API that enriches local data. This includes Literature Providers (Neurosynth, NeuroQuery) and AI Providers (OpenAI, Gemini, Hugging Face). Configuration details live in documentation/providers.

Recipe

A specific combination of CLI flags or YAML settings designed to solve a particular problem (e.g., “batch processing fMRI peaks” or “generating an anatomical report”).

Result Directory

The folder (default: coord2region-output/) where the pipeline saves artefacts. This includes the results.json, generated images, and provenance logs.

Study Radius

A spherical distance parameter (default: 10mm) used when querying literature databases. It determines how close a study’s reported peak must be to your query coordinate to be included in the analysis.

Web Runner

The client-side execution environment (hosted on Hugging Face Spaces) that allows users to run the Python pipeline directly in the browser for demonstration purposes. distinct from the Builder.