Thursday: Docking & Discussion
Overview
Thursday brings together computational concepts with practical applications and collaborative discussion. You’ll explore the differences between CPU and GPU computing through hands-on experiments, learn about molecular docking for understanding protein-ligand interactions, and participate in a roundtable discussion to reflect on the bootcamp experience.
By the end of Thursday, you’ll be able to:
- Explain what a PyTorch tensor is, what autograd does, and why every major tool this week is built on them.
- Explain why GPUs outperform CPUs for certain computational tasks, and when they don’t.
- Describe how vectorization changes the cost of numerical operations in Python.
- Describe the principles of molecular docking — sampling, scoring, and ranking.
- Apply docking concepts to protein design workflows and reason about when to use DiffDock-PP vs. PLACER vs. Rosetta.
- Articulate connections between the tools you’ve learned this week and choose the right stack for your capstone target.
Modules
| # | Module | Description |
|---|---|---|
| 1 | PyTorch Foundations | What tensors and autograd actually are, and why every ML tool this week is built on them |
| 2 | CPU vs GPU Computing | Hands-on activity comparing computational performance |
| 3 | Molecular Docking | Introduction to predicting molecular interactions |
| 4 | Reflection & Capstone Introduction | Self-reflect on your learning and plan your capstone project |
Schedule
| Time | Activity |
|---|---|
| Early morning | PyTorch Foundations |
| Morning | CPU vs GPU Activity |
| Mid-day | Docking Lesson |
| Afternoon | Reflection & Capstone Introduction |
Preparation
- Ensure you have access to Google Colab for the CPU vs GPU activity
- Review any tools or concepts from earlier in the week that you’d like to discuss
- Think about which capstone target interests you most