Bacillus Sporulation Decisions With Foldscope ML
ISEF Category: Microbiology
Ready to Turn This Idea Into a Real Project?
This guide was put together with the help of AI research tools to give you a solid starting point. But a competitive science fair project lives in the details: refining your research question, fine-tuning your variables, analyzing your data, and presenting your findings like a seasoned scientist.
For next steps tailored to your interests, skill level, and timeline, work one-on-one with a MehtA+ mentor. Learn more about MehtA+ Science & Engineering Research Mentorship →
Subcategory: Bacteriology · Difficulty: Advanced · Setup: School Lab · Time: Full Year
The Hook
A single bacterial cell can face the same stress and make a different choice from its neighbor. Some cells keep growing, and some commit to sporulation, a survival mode that acts like a tiny emergency bunker. If you can spot that split early, you can study one of the most interesting decisions in microbiology. You do not need fluorescence to start, only clear images and careful analysis.
What Is It?
Bacillus subtilis is a soil bacterium that can switch into sporulation when conditions get rough. Sporulation means the cell builds a tough dormant form that can survive heat, drying, and starvation. Think of it like a phone switching to battery-saver mode, except the cell is building a whole backup body.
The cool part is that not every cell makes the switch at the same time. That split is called asymmetry, because the population does not behave as one big uniform group. Some cells stay active, while others start the decision pathway controlled by Spo0A, a master regulator that helps turn on the sporulation program.
In this project, you would use time-lapse or repeated imaging with a Foldscope, then train a model on shape features such as size, aspect ratio, and edge texture. You would ask whether those features can predict which cells are headed toward sporulation under stress like starvation, ethanol, or heat. The big goal is to link what a cell looks like to what it is about to do.
Why This Is a Good Topic
This is a strong science fair topic because you can change one stress condition at a time and measure a clear outcome. You can compare different stresses, compare strains or media, and test whether simple image features predict a biological decision. The project connects to cell survival, antibiotic tolerance, and bacterial development. You can also build real skills in microscopy, image analysis, and machine learning without needing fluorescence.
Research Questions
- How does starvation stress change the rate at which B. subtilis cells commit to sporulation?
- How does ethanol stress change the rate at which B. subtilis cells commit to sporulation?
- How does heat stress change the rate at which B. subtilis cells commit to sporulation?
- What is the effect of cell-shape features on predicting sporulation commitment from Foldscope images?
- To what extent does a shape-based ML model match published microfluidics decision rates?
- Which image features, such as cell length, width, or curvature, best separate cells that sporulate from cells that keep growing?
- How does combining stress type and image features improve prediction of Sporulation decision rates?
Basic Materials
- Foldscope or comparable low-cost microscope
- B. subtilis strain from a school-approved source
- Sterile culture tubes and plates
- Prepared growth media approved by your teacher or lab supervisor
- Basic pipettes and sterile tips
- Disposable slides and coverslips
- Smartphone camera adapter or stable phone mount
- Digital caliper or image scale reference
- Notebook for observations and metadata
- Computer with spreadsheet software
Advanced Materials
- Phase-contrast microscope with camera attachment
- Microfluidic chips for bacterial imaging
- Temperature-controlled stage or incubator microscope setup
- Fluorescence microscope for optional validation, if available
- Image analysis workstation
- Sterile culture supplies approved by the lab
- Software for cell segmentation and tracking
- Access to published microfluidics benchmark datasets
- B. subtilis sporulation reporter strains, if the lab allows validation
- Data storage for large image sets
Software & Tools
- ImageJ: Measures cell dimensions and extracts shape features from microscopy images.
- Python: Runs image processing, feature extraction, and machine learning workflows.
- Jupyter Notebook: Organizes code, notes, and figures in one place.
- scikit-learn: Builds and tests classification models from shape features.
- PubMed: Helps you find review articles and papers on Bacillus sporulation and image-based prediction.
Experiment Steps
- Define the exact decision point you will measure, such as visible shape change, tracking outcome, or a labeled reference from published data.
- Choose one stress factor first, so you can compare a clean test case before adding more variables.
- Plan how you will image the same cells consistently enough for shape features to stay comparable across time points.
- Decide which features your model will use, then build a simple baseline before trying a more complex classifier.
- Set up controls that separate true sporulation signals from changes caused by growth slowdown, clumping, or focus drift.
- Plan how you will benchmark your results against published datasets so your model has a real comparison point.
Common Pitfalls
- Using blurry or tilted images, which makes cell-shape measurements unreliable.
- Mixing stress conditions without a clear control group, which hides which factor caused the sporulation change.
- Labeling cells only by final outcome, which can confuse early decision features with late-stage morphology.
- Letting lighting or focus shift between images, which changes shape estimates and breaks model training.
- Comparing your results to published data without matching the same imaging scale, which makes the benchmark unfair.
What Makes This Competitive
A strong version of this project does more than count cells. You would build a careful feature set, justify your classifier, and test whether it generalizes across stress types or image sources. You could also compare a simple shape-only model against a stronger baseline, then report where each one fails. That kind of analysis shows real understanding of biology, measurement limits, and machine learning.
Project Variations
- Test whether shape-based prediction works better in starvation than in heat stress.
- Compare Foldscope images from B. subtilis and another spore-forming bacterium to see whether the same features transfer.
- Add a tracking analysis that follows single cells over time instead of using only one-frame snapshots.
Learn More
- NCBI PubMed: Search for review articles on Bacillus subtilis sporulation, Spo0A, and single-cell decision making.
- NIH PubMed Central: Read full-text papers on bacterial sporulation and image-based analysis when available.
- MIT OpenCourseWare: Search for microbiology and computational biology course materials that explain bacterial regulation and data analysis.
- Bacillus subtilis and Its Closest Relatives: Look for this review title in PubMed or your school library for background on sporulation.
- ImageJ documentation: Find the official guides for measuring cell morphology and preparing microscopy images for analysis.
- NOAA and NASA image-analysis resources: Search their public guides for general tips on calibration, image quality, and reproducible measurements.