Coord2Region: Coordinates into Insights

Build a YAML config to run coord2region with minimal effort — map coordinates to region names, studies, summaries, and images.

Interactive Config Builder

Powered by JSON Schema-driven forms
How it works
  1. Choose input mode: Coordinates (paste directly or upload a CSV/TSV with x,y,z columns in MNI space) or Region names. If you upload a file, provide a valid path; for Region names, enter names directly.
  2. Select atlas(es): For Coordinates, you can pick multiple atlases (grouped by type and searchable). For Region names, select exactly one atlas.
  3. Add sources: Choose study/provider sources under the sources field. If you plan to keep Studies enabled, select at least one source.
  4. Toggle outputs: Studies are enabled by default and can be turned off. Turn on Summaries and/or Images as needed (Summaries depend on Studies).
  5. Configure Summaries (when enabled): Pick a prompt type (or write a custom prompt), select one or more summary models, and optionally set a token limit. Add provider API keys if required by your chosen models.
  6. Configure Images (when enabled): Choose image_backend (ai, nilearn, or both), select an image_model, and set a prompt type or a custom template (supports placeholders like {coordinate}, {first_paragraph}, and {atlas_context}).
  7. Save & export options: In the Outputs mini‑section, set the working directory (used for caches and downloads), choose an optional output_format, and provide an output_name when exporting.
  8. Review & copy: The YAML preview, a CLI command (that uses a saved YAML), and Direct CLI commands (no YAML needed) update live. Use the copy buttons or download the YAML.
  9. Run: Either save the YAML and run the provided CLI command (e.g., coord2region --config …) or copy a Direct CLI command matching your current selections.
  10. Tips: Reuse the same working directory to cache datasets and atlases across runs. Add provider API keys for AI features when needed.