A Visual CAD Tool for Spiking Neural Networks
Click to watch the demo on YouTube
NeuroTicks is a graphical construction kit for the NEST Simulator. It bridges the gap between abstract network definitions and spatially explicit, biological architectures.
Think of it as "CAD software for the brain": visually design, manipulate, and wire neural populations in 3D space before running simulations.
If you've tried scripting complex 3D topologies, distance-dependent connectivity rules, or procedural population distributions in pure Python — you know it gets messy fast.
NeuroTicks is built for connectomic hypothesis testing:
- How does the physical arrangement of populations affect synchronization?
- What happens to signal propagation if I twist this cortical sheet?
Note: This is a research prototype, not a replacement for PyNN. It's a prototyping assistant for spatial SNNs that lets you see what you're doing.
Define neuron geometry, not just numbers:
- Grids, Layers, Spheres
- Flow Fields (polynomial vector fields) to warp structures
- Wave Function Collapse for procedural, non-uniform distributions
Build functional modules in isolation, then combine them:
- Import existing graphs
- Use the Fusion Tool to merge networks
Watch the network spike in real-time:
- Simulate Lesions — Set connection weights to zero during runtime
- Tweak Dynamics — Adjust synaptic delays and weights on the fly
- Inject & Record — Click neurons in 3D, attach voltmeters, inject currents
Built on PyVista for high-performance rendering of neuron positions and connectivity.
Design in the GUI, export to Python + JSON, run on clusters without graphics.
Automatic snapshots of simulation runs. Reload and compare past dynamics.
wget https://raw.githubusercontent.com/Yinsalt/NeuroTicks/main/install.sh
chmod +x install.sh
./install.shThis will:
- Install all system dependencies
- Create a Python virtual environment
- Compile NEST 3.8 from source
- Clone NeuroTicks
After installation:
source ~/neuroticks/activate.sh
python Main.pydocker build -t neuroticks:latest .
./run_neuroticks_docker.shIf you prefer manual setup, ensure you have:
| Package | Version |
|---|---|
| Python | 3.10+ |
| NEST Simulator | 3.8 |
| PyQt6 | 6.x |
| PyVista | 0.46+ |
| PyQtGraph | 0.13+ |
| PyOpenGL | 3.1.10 |
| NumPy | 2.x |
| SciPy | 1.x |
| Cython | <3.0 (for NEST build) |
- Create Graph — New graph structure via menu
- Add Nodes — Define populations with spatial parameters
- Connect — Use connection tool for synaptic projections
- Attach Devices — Spike recorders, voltmeters
- Simulate — Run and observe dynamics in real-time
- Analyze — Review past runs in history tab
NeuroTicks/
├── Main.py # Entry point
├── neuron_toolbox.py # Graph/Node classes, NEST wrappers
├── WidgetLib.py # UI components, editors, analysis
├── CustomExtension.py # History browser
├── ExtraTab.py # Additional tools
└── functional_models.json
An anatomically correct retina module is under development:
- Polar projection of video input
- Spherical photoreceptor arrangement
- Correct foveal density distribution
- Direct video-to-spike conversion
- Large networks (10k+ neurons) will slow down visualization
- Tested primarily on Linux
- Documentation is minimal
- Some UI elements are rough
GPL-3.0