Circadian Splicing in Human Tissues

ISEF Category: Computational Biology and Bioinformatics

Subcategory: Genomics  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

Your cells do not run on one clock. Different tissues can follow different daily rhythms, and some genes change how they are spliced, not just how much they are expressed. That means the same gene can make different protein versions at different times of day. You can look for those timing patterns in real human data.

What Is It?

Alternative splicing is how one gene can produce more than one messenger RNA. Think of it like editing a movie in different ways. You keep the same footage, but you cut the scenes in a different order, or leave some out. That can change the protein your cells build.

Circadian regulation means a pattern that rises and falls over about 24 hours. Most people think about sleep hormones or body temperature, but splicing can also follow daily timing. In this project, you would scan large human tissue data, look for splice events that rise and fall on a daily cycle, and compare tissues. The goal is to find targets that might matter for chronotherapy, which means timing treatment to match the body clock.

Why This Is a Good Topic

This is a strong science fair topic because you can ask a clear yes-or-no question with public data. You do not need a wet lab to start, and you can test many tissues, genes, and models. The topic connects to real problems in medicine, since drug timing can affect side effects and how well treatment works. You can also learn data cleaning, statistical modeling, and how to judge false positives, which makes the project feel real and research-like.

Research Questions

• How does the strength of circadian rhythmicity in splice junction usage differ across human tissues?
• What is the effect of tissue type on the number of alternative-splicing events that pass a periodicity test?
• Does adding a periodicity-aware model identify more circadian splice events than a standard rhythm test?
• To what extent do circadian splice events overlap with genes already linked to chronotherapy or drug metabolism?
• Which tissue shows the highest fraction of rhythmic splicing events after correcting for multiple testing?
• How does sample size within a tissue affect the stability of detected circadian splice targets?

Basic Materials

• A laptop or desktop computer with enough memory to handle large tables.
• A stable internet connection for downloading public GTEx summary data.
• A spreadsheet program for tracking samples, genes, and results.
• Python installed with pandas, numpy, scipy, and statsmodels.
• R installed with tidyverse and a rhythm analysis package.
• A notebook for recording analysis decisions and version changes.

Advanced Materials

• Access to high-memory computing or a university server.
• RStudio or a command-line R environment for scripted analysis.
• Python environment with Jupyter Notebook for reproducible workflows.
• Packages for differential splicing and time-series modeling.
• A source control system such as Git for tracking code changes.
• Visualization tools such as ggplot2, seaborn, or plotly.
• Access to GTEx annotations and reference gene models.

Software & Tools

• Python: Handles data cleaning, model fitting, and custom periodicity tests.
• R: Supports statistical analysis, multiple-testing correction, and tidy plotting.
• Jupyter Notebook: Keeps code, notes, and outputs in one reproducible file.
• Git: Tracks code versions so you can compare model changes safely.
• GTEx Portal: Provides public human tissue expression data and sample metadata.

Experiment Steps

1. Define the splicing event type you will measure, such as exon skipping or junction usage.
2. Choose the tissue set and the time-related metadata you will compare.
3. Build a baseline analysis that detects rhythmic patterns in expression, then adapt it to splicing.
4. Select controls that separate true circadian signal from tissue-specific noise and batch effects.
5. Set your filtering rules, including minimum sample support and quality thresholds for splice events.
6. Plan how you will rank candidate targets for chronotherapy using effect size, periodicity, and tissue consistency.

Common Pitfalls

• Mixing tissues with very different sampling patterns, which can create fake daily rhythms.
• Treating every splice change as circadian, which inflates false positives.
• Ignoring batch effects in GTEx metadata, which can make technical noise look biological.
• Using a model that fits gene expression well but misses sparse splice-junction counts.
• Ranking targets by p-value alone, which can hide weak effects that are not useful for chronotherapy.

What Makes This Competitive

A class-level version of this project finds a few rhythmic splice events. A stronger version compares multiple rhythm models, tests whether the same result holds across tissues, and corrects carefully for multiple comparisons. You can also build a ranking score for chronotherapy that combines rhythmic strength, tissue specificity, and prior drug relevance. That kind of analysis shows real judgment, not just code running.

Project Variations

• Focus on one disease-related tissue, such as liver or brain, and test whether rhythmic splicing clusters in drug-processing genes.
• Compare a periodicity-aware model with a standard circadian method to see which one finds more stable splice targets.
• Analyze exon skipping, intron retention, or alternative donor sites separately to see which splicing mode shows the strongest daily pattern.

Learn More

• GTEx Portal: Search the GTEx site for tissue-specific expression data, sample metadata, and public documentation.
• NIH Genotype-Tissue Expression project papers: Search PubMed for review articles and methods papers on GTEx analysis.
• NCBI Gene: Look up gene annotations, transcripts, and links to related genomic resources.
• Ensembl: Use the genome browser and transcript models to map splice events to known isoforms.
• MIT OpenCourseWare: Search for free material on statistics, computational biology, and genomics analysis.
• PubMed: Search for review articles on circadian transcriptomics, alternative splicing, and chronotherapy.