A field guide to fruit fly brain hacks

What it does

This awesome-list tracks software experiments built on the Drosophila melanogaster connectome—specifically the MaleCNS and FlyWire datasets. Entries span Doom and Super Mario 64 controllers, macOS desktop pets, browser blackjack, and biomechanical simulations, each annotated with its creator, dataset, and whether it uses a full graph or a circuit subset. The list also links to research foundations, datasets, analysis libraries, and a starter template for building new fly-brain experiments.

The interesting bit

The list treats biological honesty as a feature. Every entry is reviewed and tagged with caveats like Circuit subset, Prototype, or Release pending, and the maintainers actively surface negative results—projects that document failed plasticity experiments, rewired controls, or dopamine-disconnection tests. That makes it an unusually critical survey of what connectome-driven demos can and cannot actually do.

Key highlights

• Covers roughly two dozen community projects, from Doomfly and Fly64 to DesktopFly and NeuroTerrarium, with direct links and author attribution.
• Distinguishes between MaleCNS (male central nervous system) and FlyWire/FAFB (adult female brain), and notes that attaching a simulated body does not automatically mean the brain is biologically connected to it.
• Tags entries with honest limitations: many game controllers train only a small readout layer, not the biological graph itself.
• Includes a fly-connectome-template starter with a MaleCNS soma atlas, Flybody mesh, and a React + Three.js workbench for new builders.
• Accepts community submissions via pull request and documents its review date (September 12, 2026).

Caveats

• The list is a source review, not an independent reproduction; the maintainer notes that community entries summarize their authors’ own documentation.
• Several highlighted projects are tagged Prototype, Release pending, or Circuit subset, so not everything is runnable or biologically complete out of the box.

Verdict

Worth bookmarking if you are curious about computational neuroscience, embodied AI, or the current limits of connectome-driven behavior. Skip it if you are looking for a turnkey unified fly-brain simulator—this is a catalog of individual experiments, not a framework.