AlphaFold p53 Disorder Across Mammals
ISEF Category: Biochemistry
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: Structural Biochemistry · Difficulty: Advanced · Setup: Home Setup · Time: Full Year
The Hook
Some mammals seem to resist cancer far better than others, even when their cells face the same kinds of DNA damage. p53 helps guard the genome, but parts of this protein do not hold one fixed shape. That makes it a good target for comparing structure, evolution, and disease risk across species.
What Is It?
p53 is a tumor-suppressor protein. Think of it like a security guard that checks for DNA damage and decides whether a cell should pause, repair itself, or self-destruct. In many proteins, shape is everything. For p53, some regions stay flexible instead of locking into one neat 3D form. Those flexible parts are called intrinsically disordered regions.
Your project asks whether those disordered regions look different across mammals with different lifespans and cancer rates. AlphaFold2 and ESMFold predict protein shape from sequence, and their confidence scores can hint at disorder. If a region gets a low confidence score, that often means the model thinks the area is flexible or hard to pin down. You can compare p53 orthologs, which are p53 versions from different species, and ask whether long-lived mammals show different disorder patterns than short-lived ones.
Why This Is a Good Topic
This topic works well because you can test a clear comparison question with public data and free modeling tools. It connects protein structure to aging and cancer, two real problems with huge biological impact. You can learn sequence alignment, structural prediction, confidence score interpretation, and basic statistics without needing a wet lab.
Research Questions
- How does predicted disorder in p53 vary across mammals with different lifespans?
- What is the effect of mammalian cancer incidence on pLDDT scores in conserved p53 regions?
- Does the disorder pattern in the DNA-binding domain differ between long-lived and short-lived mammals?
- To what extent do AlphaFold2 and ESMFold agree on low-confidence regions in p53 orthologs?
- Which p53 regions show the strongest link between sequence conservation and predicted disorder?
- How does the length of intrinsically disordered segments in p53 relate to species cancer resistance?
Basic Materials
- Laptop with internet access.
- Free access to AlphaFold Protein Structure Database.
- Free access to ESMFold or a similar prediction server.
- NCBI Gene and Protein pages for p53 ortholog sequences.
- UniProt entries for protein annotations and domains.
- Spreadsheet software such as Google Sheets or LibreOffice Calc.
- Sequence alignment viewer such as Jalview.
- Optional note-taking system for tracking species, scores, and source links.
Advanced Materials
- Access to a university computer cluster or GPU workstation for running local structure predictions.
- Python with Biopython, pandas, NumPy, and SciPy.
- R or Python plotting library for statistical graphics.
- Command-line alignment tools such as MAFFT or MUSCLE.
- Structural visualization software such as PyMOL or UCSF ChimeraX.
- Access to curated mammal lifespan and cancer incidence datasets.
- Local copy of species sequence files in FASTA format.
Software & Tools
- R: Makes plots and statistical comparisons for lifespan, disorder, and cancer data.
Experiment Steps
- Define the species set you will compare and decide how you will split them into long-lived and short-lived groups.
- Collect p53 ortholog sequences from trusted databases and confirm that each sequence covers the same protein regions.
- Compare sequence alignments so you can match conserved domains to predicted low-confidence regions.
- Choose one disorder metric from each model, then decide how you will summarize that metric for each species.
- Plan the cancer or lifespan dataset you will pair with each species and make sure the sources match the same taxonomic names.
- Design your statistical test and control for phylogenetic relatedness or other obvious confounders.
Common Pitfalls
- Mixing p53 isoforms or partial sequences, which makes one species look different just because the input data do not match.
- Comparing AlphaFold2 and ESMFold outputs without aligning the same residue positions, which breaks the disorder comparison.
- Treating low pLDDT as direct proof of disorder in every case, which overstates what the score can tell you.
- Using cancer incidence numbers from different databases without matching species names or study methods, which creates noisy correlations.
- Ignoring evolutionary relatedness among mammals, which can make a species group look stronger than it really is.
What Makes This Competitive
A strong version of this project goes beyond a simple species comparison. You can add a matched control set, test more than one disorder metric, and compare two prediction systems instead of trusting one. The best entries also show careful statistics, clear data cleaning, and a reasoned link between protein structure, evolution, and cancer biology.
Project Variations
- Compare p53 disorder patterns in bats, whales, and primates to see whether extreme lifespan groups cluster together.
- Swap cancer incidence for maximum lifespan and test whether structure confidence tracks longevity more strongly than disease rates.
- Focus on one p53 domain, then compare disorder changes across mammals, birds, or reptiles to broaden the evolutionary angle.
Learn More
- PubMed Central: Read free full-text papers on protein disorder, AlphaFold confidence, and comparative genomics.