C. elegans Connectome Simulation Project

ISEF Category: Computational Biology and Bioinformatics

Subcategory: Computational Neuroscience  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

A worm with just 302 neurons can still find food. That makes C. elegans a tiny but powerful test case for brain simulation. If you can model its connectome and perturb it, you can ask how simple circuits produce real behavior. This project turns a public nervous system map into a living-style computer experiment.

What Is It?

A connectome is a map of how neurons connect to each other. Think of it like a subway map for the nervous system. In C. elegans, researchers have already mapped every neuron and many of its connections, which gives you a real brain network to simulate.

Your goal is to build a model that starts with the whole network, then look for a smaller sub-circuit that still makes chemotaxis, which means movement toward or away from chemicals. Optogenetic-style perturbations mean you mimic the idea of turning neurons on or off with light, but in software. You test how behavior changes when specific neurons are silenced, boosted, or disconnected. The big idea is simple, find the smallest useful circuit that still acts like the full system.

Why This Is a Good Topic

This is a strong science fair topic because you can ask clear, testable questions with public data and code. You are not guessing, you are measuring how network changes affect simulated behavior. The real-world connection is strong too, because brain circuits, disease models, and neural prosthetics all depend on understanding how activity spreads through networks. You can learn network analysis, model fitting, simulation design, and statistics, all in one project.

Research Questions

• How does removing specific neuron classes change simulated chemotaxis accuracy?
• What is the effect of turning off inhibitory connections on chemotaxis-like movement patterns?
• Does a minimal sub-circuit preserve chemotaxis as well as the full connectome model?
• To what extent do random neuron deletions disrupt navigation compared with targeted deletions?
• Which connection strengths matter most for reproducing the chemotaxis response?
• What is the effect of adding noise to neuron activity on model stability?
• To what extent does a learned circuit generalize across different perturbation patterns?

Basic Materials

• Laptop or desktop computer with enough memory for simulation work.
• Python installed with scientific libraries.
• OpenWorm public connectome data.
• Spreadsheet software for tracking runs and outcomes.
• Basic statistics reference for comparing model outputs.
• External hard drive or cloud storage for versioned data backups.

Advanced Materials

• Workstation or lab computer with higher RAM and multi-core processing.
• Python environment with network analysis and simulation packages.
• OpenWorm data sets and curated neuron connection tables.
• GPU access if using machine learning for circuit reduction.
• Jupyter Notebook for reproducible analysis.
• Git for version control and collaboration.
• Access to a mentor familiar with computational neuroscience or graph theory.

Software & Tools

• Python: Runs the simulation, data cleaning, and analysis pipeline.
• Jupyter Notebook: Keeps code, plots, and notes in one reproducible file.
• NetworkX: Analyzes graph structure, centrality, and sub-circuit candidates.
• NumPy: Handles arrays and fast numerical operations in the model.
• pandas: Organizes connectome tables and simulation results.

Experiment Steps

1. Define the exact behavior you want to reproduce, such as chemotaxis direction, response speed, or error rate.
2. Choose one network representation, then decide how you will turn the connectome into a computable model.
3. Select a perturbation strategy that matches the biological idea of neuron activation or silencing.
4. Set up a baseline model so you can compare the full network against reduced sub-circuits.
5. Plan a reduction method that tests which neurons or edges are necessary for the behavior.
6. Predefine the metrics that will tell you whether the smaller circuit still works.

Common Pitfalls

• Using a raw connectome without cleaning edge weights, which can make the simulation reflect data artifacts instead of biology.
• Treating every neuron as equally important, which hides the effect of key sensory or motor nodes.
• Testing only one perturbation pattern, which makes the result look stronger than it really is.
• Calling any movement toward the target chemotaxis, which can blur the difference between true guidance and random drift.
• Comparing reduced circuits without a fixed baseline, which makes it hard to tell whether the simplification helped or hurt.

What Makes This Competitive

A competitive version of this project goes past one simulation run. You would compare several reduction methods, several perturbation types, and several behavioral metrics. You would also test whether your minimal circuit still works across different parameter settings, not just one lucky case. Strong projects in this area explain why the chosen sub-circuit matters biologically, not just computationally.

Project Variations

• Use a different C. elegans behavior, such as escape or locomotion, and ask whether the same reduction method still works.
• Replace the full connectome with a weighted graph version and compare how edge weighting changes circuit recovery.
• Train a classifier to predict which neurons are essential, then compare that prediction against a hand-built neuroscience rule set.

Learn More

• OpenWorm: Search the OpenWorm project site for public C. elegans data, connectome resources, and simulation background.
• WormAtlas: A free reference site for C. elegans anatomy, neurons, and nervous system maps.
• PubMed: Search for review articles on C. elegans connectomics, chemotaxis, and neural circuit modeling.
• NIH National Library of Medicine: Use NCBI Bookshelf to find free chapters on neural circuits and systems neuroscience.
• MIT OpenCourseWare: Look for courses on computational neuroscience, network analysis, and modeling.
• Nature Reviews Neuroscience: Read review articles on connectomes and circuit function through journal abstracts and accessible reviews.