Amyloid-Beta Drug Repurposing With Molecular Modeling

ISEF Category: Translational Medical Science

Subcategory: Drug Identification and Testing  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

Amyloid-beta clumps are linked to Alzheimer’s disease, and some approved drugs may stick to the same surfaces that help those clumps form. That means you can ask a real drug discovery question without inventing a brand-new molecule. Your job is to compare candidates, not guess. If you can rank drugs with careful modeling and checks against published evidence, you can build a serious research project.

What Is It?

This project asks whether existing FDA-approved drugs might bind to amyloid-beta oligomer interfaces. Amyloid-beta is a small protein fragment that can bunch into harmful clusters called oligomers. Think of an oligomer like a sticky pile of puzzle pieces. If a drug fits into the contact surface, it may interfere with the pieces coming together.

AlphaFold-Multimer predicts how proteins may interact, so you can use it to sketch possible amyloid-beta interface complexes. Then Vina rescoring estimates which candidate drugs bind more strongly to those predicted surfaces. Vina is a docking tool that gives each drug a score based on how well it fits and interacts with the target. You are not proving a drug works in patients. You are ranking candidates and testing whether the ranking lines up with known chemistry and published clues.

Literature cross-reference matters here. Some drugs may score well just because they are large, greasy, or flexible. You need to check whether the chemical features make sense and whether the literature already reports related binding, neuroprotective effects, or off-target risks. That mix of computation and reading makes the project stronger than a simple docking run.

Why This Is a Good Topic

This is a strong science fair topic because you can turn a big medical problem into a clear comparison study. You can test a set of approved drugs, measure docking scores, and compare those results against literature evidence and chemical properties. The real-world link is Alzheimer’s drug discovery, where repurposing can save time because the drugs already have safety data. You can learn molecular modeling, data filtering, and how to defend a ranking with evidence.

Research Questions

• How does the predicted binding score of FDA-approved drugs differ across amyloid-beta oligomer interface models?
• What is the effect of drug size on Vina rescoring against amyloid-beta interface pockets?
• Does drug flexibility change the ranking of top amyloid-beta binders in docking results?
• To what extent do docking scores agree with published reports of amyloid-beta binding or anti-aggregation activity?
• Which chemical features best predict favorable rescoring among approved drugs?
• How does choosing different AlphaFold-Multimer interface models change the top-ranked repurposed drugs?

Basic Materials

• Laptop or desktop computer with enough RAM for molecular modeling software.
• Access to the internet for PubChem, PubMed, and journal searches.
• Open-source molecular visualization software such as PyMOL or UCSF ChimeraX.
• AlphaFold-Multimer output files or a public protein structure source.
• AutoDock Vina or a similar free docking program.
• Spreadsheet software for tracking compounds, scores, and notes.
• Reference list of FDA-approved drugs to screen.
• Molecular file converter such as Open Babel.

Advanced Materials

• Workstation with a multi-core processor and enough storage for many docking runs.
• Python environment for batch processing and analysis.
• RDKit for calculating molecular descriptors.
• MDAnalysis or similar tools for trajectory inspection if you add dynamics.
• Molecular dynamics software such as GROMACS if your lab supports post-docking refinement.
• Access to curated drug libraries from DrugBank or PubChem.
• High-quality structural datasets for amyloid-beta oligomer models.
• Statistical analysis software for ranking comparisons and enrichment tests.

Software & Tools

• PubChem: Helps you find drug structures, synonyms, and basic chemical properties.
• PubMed: Lets you search review articles and primary papers on amyloid-beta binding and repurposed drugs.
• AutoDock Vina: Ranks candidate drugs by predicted binding to the amyloid-beta interface.
• PyMOL: Lets you inspect predicted complexes and check whether a pose makes structural sense.
• RDKit: Calculates drug descriptors such as molecular weight, ring count, and flexibility.

Experiment Steps

1. Define the exact amyloid-beta interface model you will test, and decide how many structural variants you need for fair comparison.
2. Select a screening set of FDA-approved drugs, and set rules for excluding compounds that are too large, too reactive, or poorly documented.
3. Prepare a docking workflow that treats every drug the same way, so score differences reflect the candidate, not the setup.
4. Build a scoring plan that pairs Vina results with structural checks, literature support, and simple chemical descriptors.
5. Choose controls that reveal false positives, including compounds unlikely to bind the interface and known amyloid-beta interactors if you can find them.
6. Plan your analysis before running the screen, so you know how you will rank, filter, and explain the best candidates.

Common Pitfalls

• Using only one protein model, which makes your ranking depend on a single predicted interface shape.
• Comparing docked poses without checking whether the drug actually sits on the oligomer interface.
• Treating a strong Vina score as proof of biological activity instead of a screening clue.
• Ignoring protonation state and tautomer choice, which can change how a drug docks.
• Mixing literature evidence from unrelated assays, which makes the plausibility check too loose to trust.

What Makes This Competitive

A class-level version of this project stops at a docking table. A stronger version compares several interface models, uses the same screening rules for every drug, and explains why top hits rise or fall. You can also separate drugs by chemical class, then test whether any class is overrepresented among the best scores. If you add a careful literature cross-check and a clear statistical plan, your project starts to look like a real discovery pipeline.

Project Variations

• Screen FDA-approved anti-inflammatory drugs against amyloid-beta interfaces to see whether a specific therapeutic class clusters near the top.
• Compare docking results from AlphaFold-Multimer models with results from a published amyloid-beta structure to test model sensitivity.
• Add a descriptor analysis that asks whether lipophilicity, flexibility, or aromatic ring count predicts better rescoring.

Learn More

• PubMed: Search for review articles on amyloid-beta oligomers, drug repurposing, and docking validation.
• PubChem: Find canonical structures and chemical properties for approved drugs.
• NIH Bookshelf: Read free textbook chapters on protein structure, binding, and neurodegeneration.
• NIH NCBI Structure resources: Learn how protein-ligand models are stored and viewed.
• AutoDock Vina documentation: Check the official manual and examples for docking setup and scoring.
• Nature Reviews Drug Discovery: Search the journal for review articles on drug repurposing for neurodegenerative disease.