An interactive static website for visualizing zebrafish whole-brain neural circuit inference and structure-function fusion dynamics results.
The site contains two main analysis modules:
Neural Functional Circuit InferenceVisualizes inferred directed neural networks in 3D anatomical space from zebrafish calcium imaging data.Structure-Function Fusion: Dynamics InferencePresents governing dynamical equations discovered from inferred network structure using both Symbolic Regression and Network-SINDy.
Once GitHub Pages is enabled, the site can be accessed at:
- Interactive 3D rendering of inferred neural circuits
- Edge-weight filtering for circuit visualization
- XYZ axis widget and coordinate scale panel
- Dynamics inference comparison across:
gaussian_tejoint_grangerdbn_refine
- Support for two subgraph scales:
N = 200N = 500
- Two dynamics inference methods:
Symbolic RegressionNetwork-SINDy
zebrafish-dynamics-platform/
├── index.html
├── README.md
├── .nojekyll
└── data/
├── positions.json
├── edges_gaussian_te.json
├── edges_joint_granger.json
├── edges_dbn_refine.json
├── network_dynamics_sindy.json
└── Network_dynamics_sindy.md
Run a local static server from the repository root:
cd zebrafish-dynamics-platform
python3 -m http.server 8000Then open:
Do not open index.html directly with file:// when testing data-dependent visualization, because local browser security policies may block fetch() requests to the JSON files.
Gaussian TEJoint GrangerDBN Refine
Symbolic RegressionEquation discovery from graph-informed neuronal activity featuresNetwork-SINDySparse network dynamics identification with explicit self-dynamics and coupling terms
- Frontend is implemented as a single static
index.html - 3D rendering uses
Three.js - Equation rendering uses
KaTeX - Styling uses
Tailwind CSSvia CDN
No license has been added yet. If this repository is intended for public academic reuse, a license should be chosen explicitly.