Atlas

Atlas is the actively maintained successor software for users who want to test the workflows that were migrated from neuTube, with better support for large image data and future development.

Public builds are available from the Atlas GitHub releases page. The release page provides Linux, macOS, and Windows installers, plus standalone archives. The Atlas installer also installs neuTube, so users can test the new Atlas workflows while still having the original neuTube software available. Source code and docs are available at github.com/feng-lab/atlas.

Migrated neuTube workflows

The Atlas documentation now covers the main neuTube workflows in the Atlas User Guide:

• SWC trees: load, view, edit, save, rescale, subtract, and manage neuron morphology files.
• SWC node editing: neuTube-style selection, extend, connect, break, move, radius estimation, node insertion, interpolation, type changes, and 2D/3D locating.
• Interactive tracing: trace from seed points in 2D or 3D, choose the source image/channel, and create a new SWC or attach new branches to an existing SWC.
• Automatic tracing: run the migrated neuTube auto-trace workflow as a background task, write SWC/log outputs, optionally trace on downsampled image data, and cancel long jobs safely.
• Command-line tracing: run dense in-memory tracing or blocked large-image tracing from scripts for reproducible batch workflows.
• Puncta sets: load, edit, save, detect, and export point-based annotations such as synaptic puncta.
• File formats: work with SWC, Vaa3D marker files, TIFF/OME-TIFF, Vaa3D raw stacks, HDF5/NIM, Bio-Formats-backed microscopy files, and Neuroglancer precomputed datasets.

Large-data support

Atlas was designed for larger datasets than classic neuTube. The user guide documents efficient native TIFF/OME-TIFF loading, HDF5-backed Atlas image packs, Neuroglancer precomputed datasets from local paths and remote URLs, 2D/3D synchronized viewing, and GPU-backed rendering.

For automatic tracing on very large disk-cached or Neuroglancer-backed data, Atlas adds a blocked, resumable auto-trace workflow. This mode writes a session folder with checkpoints and a final result.swc, so interrupted long runs can resume only when the same source image and tracing settings are used.

Quick test path

1. Download the latest installer for your platform from Atlas releases. The installer includes both Atlas and neuTube.
2. Launch Atlas and open an image stack, SWC file, or Neuroglancer precomputed dataset.
3. For manual tracing, enable the Trace tool, choose the image/channel and SWC target in Trace Settings, then click a seed point in the image.
4. For automatic tracing, use Image -> Auto Trace… or the Auto Trace toolbar button, choose the output SWC/log paths, and start the background task.
5. For command-line testing, use the examples in the Auto Trace CLI section.

Atlas keeps neuTube-compatible outputs such as SWC while adding modern loading, visualization, and automation paths for larger image volumes.