When I began looking at the recent fruit-fly research, what interested me was the possibility of bringing part of a neuroscience laboratory into a classroom. Researchers have made detailed maps of a fly’s nervous system publicly available, along with software that can simulate neural activity and a body moving through an environment. I would like to propose an Interactive Digital Fly Lab that brings those resources together so students can assess a biological system, form a hypothesis, intervene, and investigate the response.

This is another possible direction for a Community College Innovation Challenge team, alongside the Bedside Detective idea. Nursing, computer science, and graphic design students could contribute to a shared experimental learning environment. The specific experiments and games would be developed by the team; this initial proposal provides a starting point for faculty discussion.

The Community College Innovation Challenge, organized by AACC with the National Science Foundation, provides the context for considering this proposal. Its published rules describe a team of two to four community college students with a faculty or administrator mentor from the same institution. The team’s original work and an understandable, feasible response to a real need would matter more than the complexity of the technology. The next cycle’s requirements would guide any eventual entry; the purpose here is to look more closely at the science and what students might do with it.

The map of the nervous system. In September 2026, Berg and colleagues published a complete connectome of an adult male Drosophila melanogaster central nervous system in Cell. A connectome is a map of neurons and their connections. This one includes both the brain and the ventral nerve cord, which has a role broadly analogous to the human spinal cord. The collaboration includes HHMI’s Janelia Research Campus, the University of Cambridge, the MRC Laboratory of Molecular Biology, and Google Research. Read the paper.

The published analysis reports 166,700 fully proofread and annotated neurons, organized into 11,710 types. In comparisons with the female brain, the researchers identified 8,069 structurally corresponding types, 138 types with sex-related differences, 289 male-specific types, and 71 female-specific types. Those differences were concentrated in higher brain centers, while sensory and motor regions were largely similar. These are the counts from the published paper; the Google Research overview uses slightly different figures.

The MaleCNS project site records the release of version 1.0 on June 8, 2026, followed by the paper on September 3. It provides downloadable data, a cell-type explorer, and access to neuPrint, where users can search individual neurons and examine their connections. Neuroglancer is an open-source browser tool for viewing three-dimensional neural structures and imaging data. The MaleCNS dataset is available under CC BY 4.0, allowing reuse and adaptation with attribution, a license link, and identification of changes.

From a map to a model. An earlier milestone came in 2024, when the FlyWire consortium published the adult female brain: 139,255 neurons and approximately 50 million synapses. Dorkenwald and colleagues described that reconstruction in Nature. In the same issue, Shiu and colleagues published a computational model based on the brain’s connectivity and predicted neurotransmitter identities.

Their approach, called a leaky integrate-and-fire model, represents each neuron as accumulating incoming signals while some of that activity dissipates. When its activity reaches a threshold, the neuron produces a signal that affects connected neurons. It is a simplified account of neural activity, but it can still generate testable predictions. Stimulating modeled sugar- and water-sensing neurons predicted responses in feeding-related circuits, including neurons needed to initiate feeding. The researchers checked selected predictions in living flies; the study also examined grooming circuits.

UC Berkeley’s account of the work reports that the model can run on a laptop. That is relevant to an educational project because it establishes that useful connectome-based modeling does not always require specialized computing facilities. It does not establish that every whole-body simulation or proposed experiment would run on the same hardware.

A body for the model. A separate collaboration between Janelia and Google DeepMind produced a virtual female fly body, published online in April 2025 in Nature. Vaxenburg, Siwanowicz, Merel, and colleagues reconstructed the body from high-resolution imaging and placed it in MuJoCo, a physics simulator. The resulting model includes joints, contact with surfaces, gravity, and aerodynamic forces, allowing simulated walking and flight. Its movement controllers were trained using reinforcement learning and recordings of real flies. They were not the newly mapped male nervous system. The paper and released code and body model provide a foundation for further work.

Another body model comes from EPFL in Switzerland. NeuroMechFly, introduced in 2022 and extended in a 2024 publication, supports simulated vision, smell, and movement across challenging terrain. Its supporting Python software, FlyGym, is available under the Apache 2.0 license. The model is based on an adult female fly. That matters when considering how to combine it with male nervous-system data.

These resources are often stored on GitHub, where the code, documentation, and revision history can be examined together. Open-source licenses allow others to use and modify the software under stated conditions. A student project could therefore build on existing research tools while contributing its own educational interface, interpretation, or experiment. The data, code, images, and publications retain their respective attribution and license requirements.

What a digital fly represents. Combining a neural model with a simulated body creates a loop: sensory input enters the model, activity moves through its network, and selected outputs influence movement. Researchers and developers sometimes describe this as a digital twin. The term is useful if we also identify which features of the living animal are represented.

The recent demonstrations circulating online include game-playing and driving interfaces built around fly connectomes. One Doom project’s technical notes distinguish the measured wiring from the engineering choices that turn neural activity into gameplay. A separate Eon Systems account describes an embodied model that can forage, groom, and feed in a virtual environment, using selected descending-neuron activity to guide body-level controllers. These examples show different ways to connect a model to an environment; they are not interchangeable demonstrations of biological accuracy.

A connectome supplies anatomical connections and synapse counts, rather than a complete record of physiological connection strength, hormones, gene expression, or changing internal state. Eon’s technical discussion makes those distinctions explicit. For an educational project, the useful question is which observed responses a particular model can explain or predict. An animation alone cannot answer that question.

The Interactive Digital Fly Lab. The proposed laboratory would combine a biologically derived nervous system, a simulated fly body, and a controllable three-dimensional environment. Students would interact with that system through an accessible game-based interface. The fly would serve as the patient, the laboratory would provide its environment and assessment tools, and the games would give students ways to conduct experiments.

The central learning cycle would be assess → form a hypothesis → intervene → observe → reassess. Students would investigate a changing system, commit to a prediction, and examine the response to their intervention. The eventual team would decide which experiments and games to develop. This proposal establishes the direction for that discussion.

The educational opportunity. Nursing students practice recognizing relevant findings, interpreting them, deciding what needs attention, taking action, and evaluating the response. The Digital Fly Lab would give students another setting in which to practice that reasoning repeatedly. It could also provide opportunities for biological experimentation in courses where access to physical laboratory resources is limited. Students could manipulate variables, compare results with a baseline, and explain what they observed.

A digital patient chart. The interface would bring observations, assessments, and interventions together around the fly. Available information would come from the biological data and functions supported by the selected model. Students could move between watching the organism, examining its internal activity, and inspecting its environment.

Observation could include movement, walking, flight, feeding, grooming, rest and activity, navigation, and responses to light, odor, touch, wind, temperature, and environmental changes. A physiological view could expose modeled heart rate, cardiac rhythm, contraction, temperature, activity, and other supported measurements. A fly-specific cardiac tracing could be included where supported, rather than a human 12-lead ECG.

The proposed laboratory panel could include modeled glucose, trehalose, glycogen, triglycerides, and other lipid or metabolic measures. These would be presented in the context of fly physiology and hemolymph. An endocrine assessment could include insulin-like peptides, ecdysone, juvenile hormone, octopamine/tyramine, and other supported signaling systems. A student investigating altered feeding could compare the behavior with the organism’s modeled nutritional and endocrine state.

Neurological assessment could expose sensory input, neural activity, defined neuron populations, pathways, connectivity, and changes before and after stimulation or virtual perturbation. Imaging-style views could include the external body, internal anatomy, three-dimensional reconstruction, CT-style visualization, and neural anatomy, using the underlying biological data. The aim would be to make this information understandable without requiring a student to navigate a research database.

The environment would also be part of the assessment. Potential variables include temperature, light, odor, food, water, wind, terrain, environmental complexity, social conditions, and resource availability. Students could inspect the conditions surrounding the fly and change a supported variable to investigate its effect.

How an experiment could unfold. A student might observe that the fly detects food but does not feed. They could examine its behavior, environmental conditions, metabolic and endocrine information, sensory input, and neural activity. From that information, the student would develop a hypothesis and select an intervention supported by the model. They would predict the outcome, run the simulation, compare the response with the baseline, and decide what to investigate next.

In this interaction, “the fly is the patient” describes the learning structure. The student works through assessment, intervention, and reassessment with a computational biological organism. The result comes from running the model after the intervention, giving the student something to interpret before revising the hypothesis.

The proposed games. The games would provide different ways to interact with the same laboratory. Their mechanics would center on experiments and observable consequences. The following are concept sketches for the team to consider.

Fly Surgery would be the flagship single-player experience. A student would receive a case involving an abnormal biological state or behavior, assess the available information, and select a defined pathway or structure to investigate. An intervention could involve a virtual neural knockout or another supported manipulation. Before acting, the student would predict the result. They would then run the fly, compare behavioral or physiological changes with the baseline, and reassess. A further intervention could test a revised hypothesis. The activity would focus on what a component contributes to the system through controlled virtual experimentation.

The open FlyWire Fly Lab reports 24,250 virtual knockout experiments, including pathway knockouts, double knockouts, hub scans, and related analyses. This provides an existing example of the kind of virtual experimentation proposed for Fly Surgery, where students could investigate how a defined intervention changes the system’s response.

Neuro Detective would begin with an unexpected response. Students would use observations, sensory information, behavioral results, and neural information to develop a hypothesis. They would decide which assessment or experiment would be most useful next, run it, and interpret the result.

Pathway Priority would present several possible explanations or problems at once. Students would identify which pathway or biological process to investigate first, explain their reasoning, predict the result of the next experiment, and observe the outcome. The activity would connect prioritization with experimental design.

What Changed? would be a multiplayer comparison between a control fly or baseline run and an altered one. A neural knockout, sensory manipulation, environmental change, or other defined perturbation could produce the difference. Players would identify the change and connect the observed response to the underlying manipulation.

Fly Defender would use incoming sensory or biological information in a rapid-response activity. Students would distinguish meaningful changes from normal or background activity, using the sensory inputs supported by the laboratory.

Don’t Kill the Fly would ask students to choose an intervention for a biological problem and commit to an expected outcome before acting. The simulation might show that an intervention preserved the target behavior, disrupted another function, produced an unexpected effect, or had little measurable effect. Students would then reassess in light of the result.

Build the Behavior would let students examine the relationship between sensory input, neural processing, motor output, and behavior. They would predict what a system should do, run it, and then remove, alter, or replace a component to observe how its behavior changes.

Environmental Challenge would focus on changing one environmental variable at a time. Students could investigate supported changes in light, odor, food, temperature, wind, terrain, social conditions, or environmental complexity. Each experiment would involve a prediction, a run, a comparison with baseline, and an explanation of the result.

Classroom Fly Lab would give an instructor a shared case to present to the class. Students would assess the information, form hypotheses, choose an intervention, and commit to predictions before the instructor runs the simulation. The class could then discuss whether the outcome supported its reasoning and what to investigate next. The same system could support individual practice, laboratory assignments, live demonstrations, and group activities.

One laboratory, several educational uses. Nursing applications would center on assessment, clinical judgment, prioritization, intervention, and evaluation. Biology could emphasize experimental design, physiology, and behavior; microbiology could examine mechanisms, perturbations, and system responses. Anatomy and physiology could focus on structure, function, and regulation. Psychology and public health could explore behavior and environmental conditions. Neuroscience could examine connectivity and circuit function, while computing, AI, and engineering courses could investigate modeling, data analysis, sensorimotor systems, and embodied control.

Those are potential audiences for the laboratory. The proposed development team would still bring together nursing, computer science, and graphic design students, with the project’s specific scope and responsibilities determined through their shared work and faculty guidance.

Why the game layer matters. The proposed interaction is a cycle of predicting, manipulating, running, observing, explaining, and trying again. The learning activity would be built into what the student does to the system and how they interpret its response.

Recent research offers a reason to explore educational games. A 2026 randomized controlled trial published in Nurse Education in Practice studied 99 first-year nursing students, divided equally among serious-game training, interactive-video training, and a control group. After two weeks of training focused on parenteral medication administration, both the game and video groups performed significantly better on assessed psychomotor skills than the control group. They also reported lower stress and greater satisfaction and self-confidence in learning. The study provides a concrete example of an educational game supporting nursing students’ learning, alongside the other teaching approaches available to educators.

A pilot could examine whether students recognize relevant cues, identify meaningful changes from baseline, prioritize more accurately on later cases, improve their predictions, explain experimental outcomes in their own words, and apply what they learned. Those measures would help the team understand how the laboratory functions as a learning experience.

Access and cost. The laboratory could potentially be used in a classroom or computer lab, remotely or asynchronously, individually or with an instructor. Routine digital practice would not require animal housing or specialized physical equipment for each student session. It would provide an additional setting for experimentation alongside clinical experiences, physical laboratories, and other simulation activities.

The project would build on existing resources, including MaleCNS and FlyWire data, FlyGym, NeuroMechFly, MuJoCo, Blender, FlyBrainLab, and existing web and game-development tools. Virtual Fly Brain and FlyBrainLab are also among the resources identified in the proposal for exploring neural anatomy and circuits. The team’s work would bring selected resources together into a student-facing laboratory.

My development experience with Nurse Arcade provides another starting point. My total project spending is now approximately $1k across multiple projects and uses. That figure describes spending to date, rather than a budget for the Digital Fly Lab. Its costs would depend on the selected simulation tools, hosting, storage, and computing requirements. Reusing open resources would allow those needs to be considered in relation to the proposed laboratory.

CCIC’s judging guidance connects innovation and impact with feasibility and clarity. For this proposal, the student contribution would be an accessible experimental learning environment built from existing scientific resources. Its educational value would come from what students can investigate, the reasoning they practice, and what they learn from the results.

A student notices that something has changed, decides what information matters, makes a prediction, and watches an intervention produce a result. The next question belongs to the student: what does this tell me, and what would I investigate now?