Project Description
Antibiotic resistance is a serious real-world problem, but this project tackles one small, beginner-friendly piece of it. Students will explore how small molecules interact with a bacterial motif called D-Ala-D-Ala. Think of it as matchmaking for molecules: which candidates seem to stick, and which are an immediate no?
Using ChemGraph, guided AI prompting, and HPC, teams will inspect a starter set of molecules, try a few modifications, and run many small calculations in parallel. They will build a mini leaderboard of interaction scores, visualize their best complexes in 3D, and see why supercomputers are great at high-throughput screening.
The goal is not fake cinematic drug discovery. The goal is to learn how molecules can be represented as text and 3D objects, how batch jobs work, why embarrassingly parallel workloads scale well, and how AI can help throughout the process while judging it critically in every step.
Learning Objectives
By the end of the project, participants should be able to:
- Represent molecules as SMILES strings and 3D structures
- Run batches of independent calculations on an HPC system
- Compare approximate interaction energies to rank candidates
- Explain why this workload benefits from parallel computing
- Present scientific results and workflow lessons clearly
