Computational Protein Structure Prediction and Design

3 Credits.

The 2024 Nobel Prize in Chemistry was awarded for protein structure prediction and design. With recent research, it is now possible to blindly predict most protein structures and to design many new structures from scratch. This class will introduce the fundamental concepts in protein structure, biophysics, optimization and deep learning that have enabled the breakthroughs in computational structure prediction and design. Problems covered will include protein folding and docking, design of proteins and interfaces, and physical and deep learning methods. Class will consist of lectures and hands-on computer workshops. Students will learn to use molecular visualization tools, write programs with the PyRosetta protein structure software suite, use deep learning programs (e.g. AlphaFold2, ESM, IgLM, ProteinMPNN, RFDiffusion), review recent literature, and complete a project.