How a 9th Grader With No Research Experience Won 1st Place at His Science Fair

When Avyay Gupta came to the YRI Fellowship, he was a 9th grader with a problem most high schoolers face: he had zero research experience.

No lab access. No professor connections. No idea how to even start.

Fast forward a few months: Avyay stood at his regional science fair, presenting AI models that predict respiratory disease onset using genetic susceptibility and air pollution data. He won 1st place and qualified for state—something he never imagined possible as a freshman.

This is his story.

The Starting Point: Curiosity Without Direction

Avyay was interested in the intersection of environmental science and public health. He'd read about air pollution's effects on respiratory diseases like asthma and COPD. But he had no idea how to turn that interest into actual research.

Sound familiar? Most high schoolers face this exact problem. They have curiosity but lack:

• A research mentor to guide them
• A clear methodology
• Access to data and tools
• Knowledge of how to write and present research

Avyay's parents reached out to the YRI Fellowship looking for a solution.

The Research Question

Working with his YRI mentor, Avyay developed a sophisticated research question:

Can we integrate air pollution exposure data with genetic susceptibility markers to predict the time-to-onset of respiratory diseases like asthma and COPD?

This wasn't a simple correlation study. Avyay wanted to build predictive models that could identify at-risk individuals before they developed respiratory illness—enabling earlier intervention.

The question combined:

• Environmental science (pollution data)
• Genetics (susceptibility markers)
• Machine learning (predictive modeling)
• Public health (clinical applications)

For a 9th grader with no prior experience, this was ambitious. But that's exactly what makes research stand out to science fair judges and college admissions officers.

Building the Research Pipeline

Data Integration

Avyay's first challenge: getting data. His mentor helped him identify publicly available datasets:

• WHO Global Air Quality Guidelines for pollution exposure limits
• GEO gene expression datasets (GSE173896, GSE227691, GSE76262) for genetic markers related to COPD and asthma
• Clinical respiratory health indicators including lung function metrics (FEV1, FVC)

He learned to clean, preprocess, and merge these datasets—skills most undergraduates don't develop until junior year.

The Machine Learning Pipeline

Here's where Avyay's project became exceptional. Instead of using a single model, he implemented a multi-model survival analysis pipeline:

1. Kaplan-Meier Estimator — Non-parametric survival curves showing how genetic risk affects disease-free time
2. Cox Proportional Hazards — Semi-parametric model quantifying hazard ratios for pollution and genetic risk
3. Accelerated Failure Time (AFT) — Parametric model showing pollution causes 28% faster disease onset
4. Random Survival Forests — Ensemble method capturing non-linear interactions
5. DeepSurv Neural Network — Deep learning approach achieving highest predictive accuracy

Most PhD students don't implement five different survival analysis methods in a single project. Avyay did it in 9th grade.

Making It Explainable

One of the most impressive aspects of Avyay's research was his use of SHAP (SHapley Additive exPlanations) for model interpretability.

Black-box AI models are useless in healthcare if doctors can't understand why they make predictions. Avyay's SHAP analysis showed:

• Pollution exposure was the strongest predictor of accelerated disease onset
• Genetic risk amplified the effect of pollution (gene-environment interaction)
• Lung function (FEV1) provided protective effects

This level of explainable AI is exactly what peer reviewers and science fair judges look for.

The Results

Avyay's models produced clinically meaningful results:

The models demonstrated that combining pollution exposure with genetic markers significantly improves risk prediction—a finding with real public health implications.

Science Fair Victory

When Avyay presented his research at his regional science fair, judges were stunned.

Here was a 9th grader explaining:

• Survival analysis methodology
• Machine learning model comparison
• Gene-environment interactions
• Explainable AI techniques

He won 1st place and qualified for the state competition.

"My research won 1st place at my science fair and I qualified for state—something I never imagined achieving in 9th grade."

— Avyay G., YRI Fellowship Student

What Made the Difference

End-to-End Support

Avyay didn't do this alone. The YRI team worked with him at every step:

1. Topic Selection — His mentor helped identify a research gap at the intersection of his interests
2. Methodology Design — Structured the multi-model approach for maximum scientific rigor
3. Data Acquisition — Guided him to publicly available datasets
4. Technical Implementation — Weekly sessions debugging code and refining analysis
5. Paper Writing — Our writing staff helped draft his ISEF-format research report
6. Presentation Prep — Coached him on explaining complex concepts to judges

This is the YRI Fellowship model: comprehensive mentorship from PhD researchers who've published in top journals themselves.

Starting Early

Avyay started in 9th grade. This gave him:

• Time to iterate — Research takes longer than most students expect
• Multiple competition cycles — He can compete again in 10th, 11th, and 12th grade
• Deeper expertise — He can build on this foundation for more advanced research

If you're wondering when to start research, the answer is: as early as possible.

The Research Paper

Avyay's final paper, "Integration of Air Pollution Exposure and Genetic Susceptibility to Predict Time-to-Onset of Respiratory Disease Risk," includes:

• Abstract summarizing the multi-model approach and findings
• Literature Review grounding the work in established epidemiology (Harvard Six Cities, WHO guidelines)
• Methods detailing the five survival analysis approaches
• Results with quantitative findings and visualizations
• Discussion of clinical implications and limitations
• References to peer-reviewed sources

The paper follows ISEF format guidelines and demonstrates the scientific rigor expected for publication in peer-reviewed journals.

What's Next for Avyay

With his regional win, Avyay is now:

1. Competing at state with an even stronger presentation
2. Expanding his research to include real GWAS data
3. Targeting publication in a computational biology journal
4. Building his college application narrative around environmental health research

When he applies to colleges in a few years, admissions officers won't see just another student with good grades. They'll see a researcher who tackled a real public health problem using advanced AI methods—as a 9th grader.

That's the kind of profile that gets into Harvard, MIT, and Stanford.

Could Your Child Do This?

Avyay started with zero research experience. He didn't have:

• A family connection to professors
• Access to a university lab
• Prior coding or machine learning knowledge
• Any published papers

What he did have was:

• Curiosity about a real-world problem
• Willingness to put in the work
• A dedicated PhD mentor through YRI

The YRI Fellowship provides everything else: the mentorship, methodology, writing support, and competition coaching that transforms curious students into published researchers.

Start Your Research Journey

If you're a high school student (or parent) reading this and thinking "my child could do something like this"—you're probably right.

We've helped students with no research background:

• Publish in peer-reviewed journals
• Win science fairs
• Qualify for ISEF
• Strengthen college applications

The only requirements are curiosity and commitment.

Apply to the YRI Fellowship →

Frequently Asked Questions

Can a 9th grader really do serious research?

Yes. Avyay's project involved survival analysis, machine learning, and explainable AI—topics typically covered in graduate school. With proper mentorship, motivated high schoolers can tackle sophisticated research questions.

How long did Avyay's project take?

The core research took approximately 10 weeks of intensive work with his YRI mentor, plus additional time for science fair preparation.

What programming skills did Avyay need?

Avyay learned Python during the program. His mentor taught him pandas for data manipulation, lifelines for survival analysis, scikit-survival for Random Survival Forests, and SHAP for model interpretability.

Is this project publishable?

Yes. The research follows peer-reviewed methodological standards and is formatted for journal submission. Avyay is currently preparing it for publication in a computational biology journal.

What if my child doesn't know what topic to research?

That's normal. Most students don't know their research topic when they join YRI. Mentors help students identify research gaps aligned with their interests during the first few sessions.