Allosteric Pocket Mapping for Resistance Enzyme Targets

ISEF Category: Biochemistry

Subcategory: Medicinal Biochemistry  ·  Difficulty: Advanced  ·  Setup: Home Setup  ·  Time: Full Year

The Hook

Some antibiotic-resistance enzymes do not have just one weak spot. They can hide side pockets that may change how the enzyme works. If you can find those pockets, you can look for molecules that lock the enzyme in the wrong shape. That gives you a real drug-discovery question you can test with public data.

What Is It?

An allosteric pocket is a binding spot on a protein that sits away from the main active site. Think of the enzyme like a machine with a main button and a few side switches. If a small molecule lands on the side switch, the machine can change shape and slow down or stop.

That matters for enzymes like NDM-1 and KPC, which help bacteria break down beta-lactam antibiotics. Fpocket scans a protein structure and marks cavities that might hold a ligand. Docking then estimates how well a natural product, like a plant compound or microbial metabolite, fits in each pocket and what kind of contacts it could make.

Why This Is a Good Topic

This is a strong science fair topic because you can test clear, measurable questions with public structures and free tools. You are not guessing, you are comparing pocket size, shape, and docking score across enzymes, structures, and ligands. The project connects to antibiotic resistance, which gives your results real-world weight. You can also learn how computational drug discovery works without needing a wet lab.

Research Questions

• Which resistance enzyme, NDM-1 or KPC, has more high-scoring allosteric pockets in public structures?
• What is the effect of using different crystal structures of the same enzyme on the pocket list Fpocket produces?
• How does pocket volume relate to docking score for natural-product ligands?
• To what extent do natural products dock better to allosteric pockets than to the active site?
• Does filtering ligands by polarity and ring count improve the hit rate in predicted allosteric pockets?
• Which pocket features, such as depth or hydrophobicity, best predict repeatable docking results across structures?

Basic Materials

• Computer with at least 16 GB RAM
• Stable internet connection
• Protein structure files from the RCSB Protein Data Bank
• Natural-product ligand files from NIH PubChem
• Spreadsheet software for tracking pocket scores and docking results
• External storage or cloud backup for result files.

Advanced Materials

• Access to a Linux workstation or university cluster
• High-resolution crystal structures of resistance enzymes
• A curated natural-product library with 3D conformers
• Molecular dynamics software for follow-up stability checks
• Enzyme inhibition assay setup for validation
• Access to a plate reader or spectrophotometer.

Software & Tools

• Fpocket: Detects surface cavities and ranks them by geometric features that may support binding.
• UCSF ChimeraX: Visualizes protein structures and helps you inspect candidate pockets in 3D.
• AutoDock Vina: Estimates how strongly each natural product may bind to a chosen pocket.
• Open Babel: Converts ligand and structure files into the formats needed for docking.
• Python: Cleans result tables and makes plots that compare pockets, ligands, and enzymes.

Experiment Steps

1. Choose a small set of enzyme structures with similar quality and define your inclusion rules.
2. Run pocket detection on each structure and decide which pocket measurements you will compare.
3. Build a natural-product library and standardize every ligand file before docking.
4. Set up control dockings against the active site and against clearly weak or unrelated pockets.
5. Create a scoring plan that combines pocket geometry, docking score, and repeatability across structures.
6. Pick the plots and statistics that will show whether your pockets hold up across enzymes and ligands.

Common Pitfalls

• Comparing structures with very different resolution or missing loops, which makes one pocket look better for the wrong reason.
• Docking ligands that still contain salts, duplicate names, or bad protonation states, which skews the scores.
• Treating the top docking score as a real hit without checking whether the ligand sits in a reachable pocket.
• Mixing active-site docking with allosteric-pocket docking in the same summary table, which hides the pattern you care about.
• Using only one structure per enzyme and ignoring flexibility, which can make a temporary cavity look stable.

What Makes This Competitive

A stronger project goes beyond one docking run. You compare several structures, use clear controls, and show that the same pocket still ranks well when the protein changes shape. You can also add a decoy set, then test whether your pocket ranking beats random guessing. That kind of careful analysis makes your story much stronger than a simple screen.

Project Variations

• Compare NDM-1 and KPC pockets across several PDB structures to see which family exposes more stable remote sites.
• Swap in approved drugs or flavonoids for the natural-product library to test whether scaffold class changes docking patterns.
• Add a decoy set of matched-size compounds to check whether your pocket ranking works better than random chance.

Learn More

• RCSB Protein Data Bank: Search for NDM-1 and KPC structures, then compare resolution, mutations, and bound ligands on each entry page.
• PubMed: Search review articles on beta-lactamase allostery, docking, and natural-product inhibitors.
• NIH PubChem: Download natural-product structures, names, and property data for your ligand library.
• NCBI Bookshelf: Read free chapters on protein structure and small-molecule binding.
• MIT OpenCourseWare: Search structural biology and molecular modeling lectures for docking and protein-ligand basics.
• Frontiers in Chemistry: Search open-access reviews on allosteric inhibition and computational screening.