Web-App Overview: The LIF Connectome

This simulation visualizes a Leaky Integrate-and-Fire (LIF) spiking neural network modeled after the Drosophila brain connectome. In this model, individual neurons integrate signals from their partners based on synaptic weights and neurotransmitter identity.

How to Use

• Interactive Spiking: Click or touch anywhere on the canvas to manually trigger a sensory spike at that location.
• Behavioral Circuits: Use the buttons to trigger specific pathways:
  • Sugar Sensing: Activates Gustatory Receptor Neurons (GRNs) in the SEZ, leading to Proboscis Extension via motor neuron MN9.
  • Antennal Grooming: Triggers Johnston's Organ Neurons (JONs) which project to interneurons aBN1/2 and motor outputs aDN1/2.
• Auto-Demo: The simulation will automatically begin cycling through behaviors after 10 seconds of inactivity. Click anywhere to regain manual control.

Technical Details

The simulation utilizes the core equations derived from the 2026 FlyWire publications. The membrane potential (v) decays toward a resting potential (V_resting) of -52mV. When the threshold of -45mV is reached, a spike is emitted, and the potential resets. The connectivity weights utilize a simplified version of the Princeton Synapse update, which improved detection of axo-axonic and polyadic synapses.

Future Directions

Current whole-brain emulations are "snapshots". Future work aims to integrate neuropeptides, gap junctions, and internal states (like hunger or thirst) which are currently missing from the static EM connectome. Comparative connectomics across species (ants, bees, cockroaches) will further reveal ancestral blueprints of fundamental circuits.