Zebrafish Whole-Brain Calcium Imaging
Infer directed neural circuits from GCaMP8f volumetric calcium imaging, then discover governing dynamical equations that describe network-level activity propagation.
Upload calcium imaging data and select a causal inference method to reconstruct the directed neural network in 3D anatomical space.
Upload the activity dataset, choose an inferred functional network, and identify the governing equation for network-level activity propagation.
Upload calcium imaging data and select a causal inference method to reconstruct the directed neural network.
Step 1 — Input Data
Click or drop a MATLAB whole-brain activity file
Required: .mat · MATLAB 7.3 / HDF5 container · validated locally before submission
Uploading MATLAB dataset...
Transferring file to the analysis workspace over a managed upload channel
This transfer may take a few minutes.
MAT file validated successfully
Ready for circuit inference
File
—
Size
—
cell_response
—
Validated Keys
—
Step 2 — Select Inference Method
Gaussian Transfer Entropy approximation. Measures directed information flow between neuron pairs using lag-1 conditional mutual information.
Joint-lag pairwise Granger causality. Tests whether past activity of neuron i improves prediction of neuron j beyond its own history.
Discrete first-order Dynamic Bayesian Network refinement on candidate edges. Produces the sparsest and most selective directed graph.
Edge Weight Filter
Only render edges whose weight is greater than or equal to the selected threshold.
X Extent
—
Y Extent
—
Z Extent
—
Normalization
—
Start from the uploaded whole-brain activity dataset, then choose an inferred network structure and subgraph scale to identify the governing dynamical equation for activity propagation.
Step 1 — Input Data
Validated activity dataset
Upload the MATLAB activity file before running the dynamics module.
File
—
Size
—
cell_response
—
Container
MATLAB 7.3
Step 2 — Select Network Structure
Select the circuit inference family used to derive the functional network from the uploaded activity dataset, then continue with dynamics identification on that network.
Step 3 — Select Dynamics Inference Method
Searches over candidate symbolic terms and selects the equation that best predicts temporal activity change from the inferred network and neuronal activity features.
Sparse network dynamics identification with explicit self-dynamics and coupling terms. STLSQ is the sparse thresholded least-squares solver used inside this method, not a separate biological model.
Step 4 — Select Informative Subgraph Scale
Reward
—
MSE
—
Complexity
—
Interpretation
Variable Definitions