Natural SGLT2 Inhibitor Screening

ISEF Category: Biomedical and Health Sciences

Subcategory: Nutrition and Natural Products  ·  Difficulty: Advanced  ·  Setup: Home Setup  ·  Time: 1 to 2 Months

The Hook

Some plant molecules may fit a diabetes target better than you expect. SGLT2 is a protein that helps move sugar back into the body, so blocking it can lower blood glucose. Your project can test which natural compounds look like good keys for that lock. Then you can check whether the fit stays steady in molecular dynamics, not just in a one-shot docking score.

What Is It?

Molecular docking is a computer test that asks which small molecule can sit best in a protein's binding site. Think of the protein as a glove and each phytochemical, a plant-made chemical, as a hand shape. If the shape lines up well, the score improves.

In this project, you screen phytochemicals from IMPPAT and AfroDB against SGLT2, short for sodium-glucose cotransporter 2, a kidney protein that helps pull glucose back into the blood. After docking, molecular dynamics checks whether the best poses stay put while the protein and water move around. That gives you a stronger clue than docking alone, because real proteins are not rigid.

Why This Is a Good Topic

This topic works well because you can measure it with clear scores, compare many compounds in a fair way, and connect the results to a real health problem, type 2 diabetes. You can learn how to clean chemical data, rank docked poses, and test whether the top hits stay stable in MD. A strong project can also compare two plant databases with the same rules, which gives you a real research question instead of a simple search.

Research Questions

• Which phytochemicals from IMPPAT and AfroDB produce the strongest docking scores against SGLT2?
• How does the ranking of top hits change when you compare Indian and African phytochemical libraries with the same receptor prep?
• What is the effect of adding a known SGLT2 inhibitor as a control on how you judge the plant compounds?
• Does molecular dynamics keep the top docked phytochemicals in the SGLT2 binding pocket more steadily than weaker hits?
• To what extent do hydrogen bonds and hydrophobic contacts persist during MD for the best-ranked compounds?
• Which compound class, such as flavonoids, alkaloids, or terpenoids, gives the most stable SGLT2 binders?

Basic Materials

• Laptop with at least 8 GB RAM and stable internet access.
• Google Colab account for running notebooks in the cloud.
• Spreadsheet software such as Google Sheets or Excel for tracking scores.
• Free molecule viewer such as PyMOL, ChimeraX, or UCSF Chimera for inspecting poses.
• Compound lists from IMPPAT and AfroDB in downloadable file format.
• PubChem access for checking structures, synonyms, and basic properties.

Advanced Materials

• Linux workstation with a modern CPU and GPU support for longer simulations.
• University cluster or server account for repeated molecular dynamics runs.
• Protein preparation software and trajectory analysis tools for receptor setup.
• Fast storage drive for docking grids, trajectories, and output files.
• Curated structure files for SGLT2 and known control ligands.

Software & Tools

• Google Colab: Runs docking and molecular dynamics notebooks without local hardware limits.
• AutoDock Vina: Scores how well each phytochemical may fit the SGLT2 binding site.
• RDKit: Cleans structures, removes duplicates, and calculates simple compound filters.
• PyMOL: Shows docking poses and binding contacts so you can compare hits by eye.

Experiment Steps

1. Define one receptor structure, one control set, and one scoring rule so every compound gets judged the same way.
2. Build and clean your ligand library, then separate the compounds by source database and chemical class.
3. Set up docking conditions that let you rank hits, inspect poses, and remove obvious false positives.
4. Choose a small top-hits set for molecular dynamics, and decide which stability metrics will count most.
5. Combine docking score, contact persistence, and trajectory stability into a final shortlist that answers your research question.

Common Pitfalls

• Preparing the protein differently for each batch, which changes the binding pocket and breaks score comparisons.
• Using compounds from IMPPAT and AfroDB without deduplicating them, which makes the same molecule look like two separate hits.
• Trusting the top docking score without checking whether the pose points into the active site.
• Skipping a known control inhibitor, which leaves you with no baseline for judging how strong the plant hits really are.
• Reading molecular dynamics from one snapshot instead of the full trajectory, which hides unstable binding and noisy outliers.

What Makes This Competitive

A stronger version of this project does more than rank the top scores. It uses the same receptor prep, the same control set, and the same filters for both databases, then checks whether the best hits stay stable in molecular dynamics. You can also compare chemical classes or run repeats to see whether the pattern holds. That kind of design gives you cleaner evidence and a better story than a simple docking list.

Project Variations

• Screen only one chemical family, such as flavonoids or alkaloids, to see whether a single class dominates the best SGLT2 hits.
• Replace SGLT2 with another diabetes target, such as DPP-4 or alpha-glucosidase, to compare where the same phytochemicals bind best.
• Add basic drug-likeness filters before MD to test whether the strongest binders also look better on paper.

Learn More

• PubMed: Search review articles on SGLT2, natural products, docking, and diabetes pharmacology.
• PubChem: Find compound structures, synonyms, and downloadable files for ligands and controls.
• IMPPAT: Browse Indian medicinal plant compounds and filter by chemical class.
• AfroDB: Browse African natural products and collect structure files for screening.
• NCBI Bookshelf: Read free background chapters on glucose transport, protein structure, and diabetes.
• MIT OpenCourseWare: Search free course notes on biochemistry, medicinal chemistry, and molecular modeling.