A curated collection of fruit fly connectome projects, from Doom and desktop pets to brain models, biomechanical bodies, and research tools.

Explore what people are building with MaleCNS, FlyWire, and the Drosophila melanogaster connectome. Each entry links to its creators' repository and explains what it actually offers.

Start with a game, explore a desktop fly, or build your own using the research foundations.

Built something with a fly connectome? Add your own repository with a pull request. Read the contribution guide, open a pull request, or suggest a project.

• Start here
• Games and control experiments
• Desktop flies and interactive worlds
• Language, art, and other experiments
• Brain models and embodied simulation
• Datasets and official resources
• Analysis libraries and viewers
• Tutorials and papers
• Related lists
• Contributing

A connectome records neurons and their connections. A simulation adds assumptions about neuron dynamics, sensory input, and motor output. A moving fly, changing weights, or a game demo does not by itself demonstrate biological fidelity or learned behavior.

• MaleCNS covers the adult male central nervous system, including brain and ventral nerve cord. The official project dates v1.0 to June 8, 2026 and the Cell paper to September 3, 2026. Project and release history.
• FlyWire / FAFB is an adult female brain connectome, a different dataset from MaleCNS. FlyWire.
• Flybody and NeuroMechFly supply simulated bodies and environments. Using either does not automatically connect a brain model to the body.

Reading the list: community entries summarize their authors' documentation, reviewed on September 12, 2026. These are source reviews, not independent reproductions. Prototype, circuit subset, and release pending identify useful limitations. Neuron counts depend on the release and filtering; synaptic contacts and directed neuron-pair connections are different quantities.

• Doomfly by nftechie - MaleCNS simulation connected to ViZDoom through modeled visual inputs and a fixed button readout. Includes plasticity experiments and negative validation results; learned survival has not been demonstrated.
• FlyDoom by eganeganegan - Research framework comparing a MaleCNS-constrained sparse recurrent controller with rewired graphs, conventional neural networks, and other controls on simple VizDoom tasks.
• Fly64 by ornata - Hooks a MaleCNS model up to Super Mario 64, with a local dashboard and macOS setup. A playful experimental controller; requires your own game ROM.
• Fly Brain Minecraft by blendi-remade - Fabric mod that runs a filtered MaleCNS graph for each fly mob, with modeled sensory inputs, motor readouts, and live neural HUDs.
• fly-flappy by ns2250225 - Browser Flappy Bird experiment running the full MaleCNS v1.0 graph at 50 Hz with engineered visual inputs and either raw DNp01 decoding or a trainable logistic action readout. Full retained graph; interface and readout are artificial. The source repository currently has no separate license file.
• Flyhard by MarkUnthank - Trains a MaleCNS-based model to operate a steering wheel through simulated fly limbs and connect it to CARLA. Documents a bounded steering result; visual driving remains a future milestone.
• Help the Fly Escape by dzhng - Arrange household objects and observe flies navigating a 3D browser game. Rust/WASM simulation using a selected MaleCNS circuit. Play.
• Swat by hrook1 - Browser arcade game with a visible escape circuit: 6,000 MaleCNS neurons influence evasive movement alongside authored game mechanics. Circuit subset. Play.
• FlyBrain by Jhongdlp - MaleCNS-based game-boss experiment with a Rust neural engine, Python training environment, and Three.js viewer.
• FlyPong by jonatasperaza - Pong controller built from a MaleCNS subgraph, with documented controls and negative results for its dopamine-inspired plasticity experiments. Circuit subset.
• Connectome Fighter by Unjuno - MaleCNS controller connected to FightingICE through explicit sensory and action mappings, with a research ledger separating connectome structure from project-defined learning rules.
• Fly Chess Lab by tolatolatop - FlyWire chess experiment with a Rust/WASM LIF simulation, spike traces, and disconnection controls. Uses an engineered, currently untrained move readout. Demo.
• fly-craftax by liuzihe02 - Connectome simulation connected to Craftax, with PPO training of a descending-neuron readout, baseline comparisons, and a replay viewer. Research prototype.
• Fly Dino by cobanov - Original Chromium Dino controlled by a fixed 80-neuron MaleCNS circuit with simplified leaky tanh dynamics and a 243-parameter readout trained with the cross-entropy method (CEM). Includes live computed activity, reproducible training, held-out benchmarks, and circuit-silencing controls. Circuit subset: only the artificial readout is trained. Play · Methods and results.
• NeuroCraft Fly by evnsnclr - Minecraft project with a recorded MaleCNS-based interactive demo and a release roadmap. Release pending: the public repository currently contains project materials, not runnable mod or companion source.
• FLYFEAR by furkancak1r - Godot horror prototype where a MaleCNS-derived simulation influences a fly and an external adaptive event policy. Includes training and validation reports. Play.
• FLYT3 by seanphan - Tic-tac-toe against a full-graph MaleCNS LIF simulation with a REINFORCE-trained descending and VNC readout, plus a fire-watch mode that classifies satellite wildfire-damage tiles through the same frozen circuit (60.8% on 5 classes) with live activity and region-lesioning views. Full retained graph; only the readout is trained. Reports its own negative control: a pooled-pixel MLP on the same inputs scores 85.7%. Self-hosted demo, no public link.
• fly-swing by JHC56 - A MuJoCo fruit fly that web-swings through a random obstacle course like Spider-Man. There is no dodge code. The escape is fired by the real giant fiber circuit from MaleCNS (LC4/LPLC2 -> DNp01, LIF), and the whole brain runs live on screen. The fly picks where to swing with a kNN threat memory built on whitetree. It reaches the goal on 20 out of 20 random courses. Circuit subset drives the fly; full retained graph is display only.
• The Fly's Table by WilliamJones - Heads-up blackjack against a MaleCNS mushroom-body model: 682 projection neurons, 1,200 Kenyon cells, 97 output neurons and 332 dopaminergic neurons with measured wiring, learning at the table through a project-defined dopamine-gated depression rule. One seat, a queue, spectators, and a brain shared through Firestore that persists between players. The same repository holds an 80-neuron Dino controller, a Flybody walker, and odour-conditioning experiments, each with silenced-circuit or dopamine-cut controls. Circuit subset; only Kenyon-cell-to-output synaptic gains change. Reports its own results: 30,000 solo hands move the fly from a -45% to a -16% winrate with 52% agreement with basic strategy, disconnecting dopamine removes the gain, and a poker attempt is documented as a negative result. Play.
• Fly-NAF by ArtyMend07 - Five Nights at Freddy's 1 played by the whole FlyWire v783 graph, 138,639 LIF neurons with their measured weights and nothing trained in the loop. DNp01 slams the doors, accumulated DNp09 potential raises the camera tablet, and raising it drives the GABAergic inhibitor clusters so the fly goes blind while on cameras. Builds on fly-brain, GPL v3. Full retained graph; no trained readout. A starvation timer forces a look at any hallway unwatched for thirty seconds, counted separately in the session report. Windows only, no playable demo. Best run reached 4 AM on night 2. Recording.
• Fly Self Driving by suanmiao - Drives a simulated street from a 64x32 windscreen using the MaleCNS connectome (165,122 neurons, 25,563,197 synapses). Complete retained graph: the wiring is frozen; only one gain per synapse and one leak per neuron are learned, with a random frozen pixel map and steering readout, and the network state is carried between decisions. Completes 58 of 60 across three held-out sets with no car collisions, against a randomly rewired graph (16/20), a linear map (12/20) and an MLP (18/20). One training seed per condition; it does not brake yet. Demo
• Banana Quest by professorwang - Browser foraging game on the full FlyWire FAFB v783 graph: 139,255 LIF neurons run in a Web Worker, odor-gradient input is split across left/right antennal ORN pools, and turning, feeding, and startle-escape are gated by descending-neuron (GNG_DESC) and SEZ/motor readout, with a live 63-group activity view and an in-app honesty panel separating measured wiring from artificial mappings. Zero-dependency vanilla JS, reusing the snedea/flybrain data packaging. Full retained graph; sensory mapping and readout pools are artificial. Play.
• fly-cartpole by Curt-Park - Balances CartPole-v1 as a reflex with the MaleCNS flight-stabilisation circuit: 5,459 neurons from the haltere, ocellar and HS-cell inputs to the wing steering motor neurons. Circuit subset; the wiring stays frozen and only five sensory input gains are tuned, with a reward signal computed outside the circuit. Untuned it lasts 275.6 steps (random pushes: 22.1); self-tuned it reaches 499.9 on held-out seeds. Demo.
• Doodle Fly by dtecx - Doodle-Jump-style browser game steered by the whole FlyWire v783 brain: 138,639 LIF neurons with the Shiu et al. 2024 model and parameters run in a Web Worker; the target platform drives LC10a on one side, and six descending-neuron types (incl. DNa02), right minus left, press the left and right buttons after a start-up calibration. The sugar -> proboscis and looming -> giant-fibre circuits are live too. Full retained graph; target selection, LC10a drive and the button readout are hand-made; nothing is trained. Reports controls: 0-1 falls in 2 minutes intact, 41-44 with the eyes swapped, hopping in place when blind. Play · Results.
• 20fly8 by dtecx - 2048 learned in the browser by a FlyWire v783 mushroom-body model: each board cell and tile size drives its own projection neuron (32 of 267), the measured PN -> Kenyon-cell (27,502) and Kenyon-cell -> MBON (62,261) connections carry the smell to 96 output neurons, and approach vs avoidance is read from which dopamine cluster (PAM or PPL1) innervates each MBON. A project-defined dopamine rule driven by a reward-prediction error (TD learning on afterstates) changes only the Kenyon-cell -> MBON gains. Circuit subset; firing-rate model, not spikes. Reports frozen-synapse controls over 500 games: random moves 1,103, naive fly 1,594, trained after 3,000 games 6,986 (512 in 75%, 1024 in 15%), nose rewired 925, synapses shuffled 1,287. Play · Results.

• DesktopFly by DenisSergeevitch - A macOS desktop fly combining FlyWire spiking circuits with a MaleCNS brain-to-leg extract, modeled senses, and articulated behavior. Includes an Electron port.
• gnat by lubabs770 - Linux/Hyprland desktop port of DesktopFly with a brain window and stimulation controls. Uses a 668-neuron circuit subset.
• DesktopFly for Linux by somsom10 - Python/GTK port of DesktopFly for GNOME on X11 and Wayland, retaining the upstream FlyWire circuit and documenting platform limitations.
• FlyBrain by snedea - Browser-based FlyWire FAFB v783 LIF simulation with food, touch, light, and temperature inputs and a live neural activity display.
• FlyWire Neuro by pusulamkendim - Local FlyWire LIF simulation linked to a persistent 3D body, sensory controls, and recorded runs. Descending-neuron activity selects measured or cached motor behavior. Research prototype.
• Closed-Loop Fly by ZeroXClem - Browser sensorimotor loop joining a modeled compound eye and optic lobe to MaleCNS descending-neuron readouts and a simulated flying body.
• flyverse by tel-0s - MaleCNS simulation with modeled color vision, smell, taste, wind, locomotion, and an inspectable room environment. Sensory and motor physiology remain project assumptions.
• NeuroTerrarium by 5p00kyy - Inspectable 3D lab with a synthetic terrarium and a separate 51-body MaleCNS giant-fiber microcircuit. Reports a negative topology-control result. Circuit subset.
• FLYBOARD by NullLabTests - CPU arcade interface for injecting current into named MaleCNS populations while viewing the modeled voltage and spike field.
• Infinite Sugar by cnqso - Browser artwork placing a FlyWire-based fly in a terrarium with continuous sweet-sensing input. Neural activity drives some movements while wings and small foot motions use supplied patterns. Experience.

• FLM by nftechie - Frozen language model coupled to the retained MaleCNS graph through a trained readout adapter. Language ability comes from the pretrained language model.
• fly-brain by lixiang1076 - FlyWire LIF simulation with a natural-language interface and project-defined dopamine-learning experiments.
• Fly / Wirehead by mattyhempstead - A MaleCNS simulation receives frames from insect videos on a virtual phone. Local Python/C++ simulation, browser observation chamber, and neural telemetry.
• FlyScroll by ranagwho - MaleCNS-based video-feed experiment where modeled visual and mushroom-body activity influences when the viewer advances to the next clip.
• Stonkfly by nftechie - Experimental MaleCNS controller with market-chart inputs, paper trading, and optional Coinbase integration. Includes modeled reinforcement and memory; profitable learning has not been demonstrated.
• OpenFly by marketcalls - MaleCNS-based controller for a paper-traded NIFTY options strategy through OpenAlgo. Live trading is opt-in and no profitable edge has been demonstrated.
• Fly Lab by Apolotary - Ableton Live music experiments driven by simulated fly motor circuits and a small trainable musical readout. The measured wiring remains fixed.
• Faiku by xyzzyapps - Haiku and glyph-tracing experiment using MaleCNS simulation and mushroom-body-inspired reinforcement, with a separate reduced fallback model.
• Fruitless by nicodunks - MaleCNS intervention study and Three.js visualization testing how blocking modeled mAL output changes activity in courtship-related candidate neurons.
• fly-odor-onoff by yukincom - Reproducible odor ON/OFF intervention experiments using Fly Brain Minecraft's filtered MaleCNS LIF graph. Includes input-pathway audits and four-PN DM1/VA2 responses under externally scheduled OFF-only input blocking. Fixed circuit; no biological validation. Methods and results.
• Fruit Fly Fashion by jtc268 - Reproducible art experiment using MaleCNS spike vectors to control the placement, rotation, and scale of print designs.
• mindmeld-with-fly by Decentricity - Sparse MaleCNS reservoir experiments with terminal rendering and recording/replay. Prototype: the README places the EEG interface in a future phase.
• boltzmann-fly by jniimi - Energy-based world model of simulated consumer behaviour (the author's own Purchase World model, ICONIP 2026) re-trained as a Boltzmann machine whose couplings are masked to an extracted MaleCNS circuit: the right-hemisphere mushroom body (PN -> Kenyon cell -> MBON). Ships the original dense model as an anchor, degree-preserving and Erdős–Rényi rewiring controls, CPU-reproducible weights, and a results note. Research prototype: the fly wiring matches the original on prediction and on one counterfactual parameter, is below it on the other, and flips the sign of the free-energy clamp test (one seed).
• BioReservoir by igdigitallab - Whole retained MaleCNS graph run as a fixed reservoir for visitor questions: a text embedding sets Poisson rates on real sensory populations, the full brain runs 250 ms of LIF per trial, and the answer is the lateral bias across descending neurons after subtracting each brain's intrinsic handedness measured on neutral reference sentences. Reproduces Shiu et al.'s sugar -> MN9 response; the female BANC connectome fails the equivalent lateral check and that failure is published rather than dropped. Not trained on outcomes and makes no forecasting claim - entertainment and research. Per-run measurements against rewired-connectome, no-connectome and random-graph controls are published in the repository. Live page · results. Submitted by the project's own team.

• Drosophila brain model by philshiu and collaborators - Research code for the Shiu et al. connectome-based leaky integrate-and-fire model, including activation/silencing experiments, notebooks, and FlyWire data configuration.
• fly-brain by Eon Systems - FlyWire whole-brain LIF implementation based on Shiu et al., with multiple simulation backends and benchmarking tools. This repository supplies the neural model; it is not a complete embodied demo package.
• flybody by TuragaLab / Google DeepMind / HHMI Janelia - Anatomically detailed MuJoCo fruit-fly body, walking and flight environments, and reinforcement-learning examples. A body and control platform.
• FlyGym / NeuroMechFly by NeLy-EPFL - Python framework for embodied sensorimotor experiments with a biomechanical fly, sensory interfaces, and physical environments. Check version-specific documentation when following older tutorials.
• flyvis by TuragaLab - PyTorch implementation of connectome-constrained models of the fly visual system, with pretrained models and analysis tutorials.
• train-your-fly / connectome by eudald-seeslab - PyTorch Geometric toolkit for connectome-constrained vision models: a Drosophila eye model feeding message passing over the FlyWire v783 whole-brain graph to a Kenyon-cell readout. Fixed wiring; only per-synapse gains and a linear readout are trained. The connectome study adds colour, shape, and numerosity tasks, randomized-ensemble controls, and manifold analysis; graphs archived on Zenodo.
• webgpu-fly by abgnydn - Browser-based WebGPU/WASM experiment joining FlyWire brain activity, MANC nerve-cord data, and a Flybody body through explicit approximate motor mappings. Demo.
• NeuroFly by seven-monarchs - Closed-loop personal research project coupling FlyWire-based neural activity to a NeuroMechFly body, with recorded simulations and documented assumptions.
• Fruit Fly Lab by vaibhavkedarisetti - Interactive FlyWire v783 whole-brain LIF simulation with stimulus controls and activity visualization. Research prototype.
• CHIMERA by caparison1234 - Drosophila larval connectome experiment linking a 1,373-neuron circuit to a MuJoCo body and a language-to-stimulus interface. Larval circuit subset.
• Connectome OS by ruvnet - Rust LIF runtime and debugging interface for stimulating, cutting, and measuring FlyWire-derived graph activity. Alpha research preview.
• FastFly by eonfathom - CUDA and CuPy simulator targeting real-time or faster execution of the FlyWire v783 graph on a consumer NVIDIA GPU.
• MaleCNS on Apple MPS by seohyunjun - Experimental PyTorch MPS simulator for Apple Silicon with verified data preparation, named-cell stimulation, and 3D activity reports.
• drosophila-brain-mlx by Kisame76 - Apple MLX and Metal port of the Shiu et al. LIF model on the complete FlyWire v630 graph, with a MaleCNS v1.0 pack, upstream-compatible run_exp output, and validation against the Brian2 reference. Includes a degree-preserving shuffled-connectome control. Apple Silicon only. Control result.
• AxonWeave by dhakalnirajan - Python library exposing MaleCNS as a sparse, trainable substrate for NumPy, PyTorch, and TensorFlow while keeping source topology and modeling policies explicit.
• Embodied fly-brain by erojasoficial-byte - Community research repository combining a FlyWire spiking model with NeuroMechFly/MuJoCo, sensory experiments, and an accompanying preprint.
• Wired Different by dhruvin-sarkar - classifier predicting sexually dimorphic cell types from male CNS connectome (male-cns v1.0) wiring, validated against degree-preserving null graphs; static demo with atlas, pathfinder and guessing game.

• MaleCNS - Official male CNS project, cell-type exploration, release history, and downloads. Collaboration between FlyEM/HHMI Janelia, Cambridge, MRC LMB, and Google Research.
• MaleCNS project source - Source for the project website and Dimorphism Explorer.
• MaleCNS Cell Type Explorer - Source and generated documentation for browsing MaleCNS v1.0 cell types, connectivity, regions, and morphology.
• MaleCNS supplemental data - Derived data and analysis notebooks accompanying Berg et al., including sensorimotor flow and cell-type information.
• Male optic lobe connectome code - Analysis and reproduction code accompanying the complete male visual-system connectome study.
• BANC project - Data, figures, notebooks, and documentation for the adult female Brain-And-Nerve-Cord connectome and its accompanying publication.
• FlyWire - Adult female brain reconstruction, research resources, and links to exploration tools.
• FlyWire annotations - Neuron annotations and other data products for the FlyWire v783 release.
• Drosophila neurotransmitters - Versioned literature-derived neurotransmitter ground truth linked across major fly connectome datasets.
• Optic lobe annotations - Cross-dataset cell-type matches between FAFB/FlyWire and the MaleCNS optic lobe.

Follow each upstream dataset's citation and licensing requirements. The license of this list does not relicense linked code, datasets, or artwork.

• Codex by Murthy Lab - Source for the FlyWire Connectome Data Explorer. Explore neurons.
• FlyBrainLab - Interactive platform for exploring fly brain data, constructing executable circuits, and studying their modeled function.
• EOScircuits by FlyBrainLab - Executable models of the Drosophila antenna, antennal lobe, and mushroom body for early olfactory-system experiments.
• NAVis by navis-org - Python tools for neuron morphology analysis, visualization, transformations, and data access.
• fafbseg by navis-org - FlyWire/FAFB segmentation tools for meshes, skeletons, annotations, and connectivity queries, interoperable with NAVis.
• navis-flybrains by navis-org - Template brains and transforms for mapping neurons between Drosophila datasets and coordinate spaces.
• connectome-interpreter by YijieYin - Python library for effective connectivity, path finding, circuit manipulation, and differentiable models at whole-brain scale.
• connectome data prep by YijieYin - Prepared connectivity matrices and reproducible preprocessing for MaleCNS, BANC, FlyWire, hemibrain, and other datasets.
• cocoa by flyconnectome - Python library for comparative connectomics, co-clustering, cell typing, and matching across FlyWire, hemibrain, MANC, and MaleCNS.
• Connecto by schlegelp - Unified Python interface that normalizes connectivity, metadata, and morphology queries across CAVE and neuPrint datasets.
• DROCAT by Swida-Alba - Python and web toolkit for path finding, network visualization, morphology, and cross-dataset analysis across neuPrint, FAFB, and BANC.
• BigClust 2 by flyconnectome - Interactive application for exploring high-dimensional connectomic clusterings with linked 2D, 3D, connectivity, and annotation views.
• neuVid by connectome-neuprint - Generates anatomical Drosophila videos from high-level descriptions using neuPrint data and Blender or VVDViewer.
• malecns by natverse - R access to MaleCNS data with metadata and morphology conveniences.
• coconatfly by natverse - Comparative connectomics across fly datasets, including FlyWire, hemibrain, MANC, FANC, and MaleCNS.
• fafbseg for R by natverse - R tools for FlyWire/FAFB segmentation, meshes, annotations, and connectivity analysis.
• neuprintr by natverse - R client for neuPrint, designed to work with the wider natverse neuroanatomy ecosystem.
• bancr by natverse - R package for querying BANC metadata, connectivity, skeletons, and meshes.
• neuprint-python by connectome-neuprint - Python client for querying connectivity and metadata through neuPrint.
• CAVEclient by CAVEconnectome - Python client for the Connectome Annotation Versioning Engine, used to access versioned connectomic data and annotations.
• FlyWire network analysis by Murthy Lab - Research scripts and notebooks for network statistics, motifs, connectivity, and graph structure.

• Fly connectome data tutorial - Workshop materials for loading, analyzing, and visualizing major fly connectome datasets using Python and R.
• FlyWire data access - Seung Lab tutorials for programmatic access to FlyWire data, meshes, and annotations.
• Hemibrain examples - Code examples accompanying Schlegel and Bates et al. for exploring cell types and circuits in the hemibrain dataset.
• Neuronal wiring diagram of an adult brain - Dorkenwald et al., Nature (2024), the adult female brain connectome.
• Whole-brain annotation and multi-connectome cell typing of Drosophila - Schlegel et al., Nature (2024), cell annotations and comparisons across connectomes.
• Connectome-constrained networks predict neural activity across the fly visual system - Lappalainen et al., Nature (2024), the research behind flyvis.
• Whole-body simulation of realistic fruit fly locomotion with deep reinforcement learning - The research accompanying flybody.

• Awesome Fruit Fly Connectome - An earlier community list of datasets, tools, papers, and fly-connectome experiments.
• Fly Connectomics tools - Overview of R and Python neuroscience tools maintained by the Cambridge Fly Connectomics Group and Jefferis Lab.

Small projects are welcome. We prioritize a clear connection to fly connectomics, useful documentation, original creators, and accurate descriptions over star counts. Substantial ports belong here when their upstream origin is credited. Announcement-only projects must say that runnable code is pending.

Built a project that belongs here? Read CONTRIBUTING.md, add your repository to the relevant section, and open a pull request. You can also use the project suggestion form if you do not want to edit the list yourself. The list is maintained by cobanov, who also maintains Fly Dino.

CC0 1.0 Universal. Linked projects retain their own licenses.