Computational Protein Design

Learning Outcomes

This module introduces the structural and energetic principles underlying protein architecture, forming the basis for computational design. It covers energy functions and scoring models used to evaluate protein conformations, along with key algorithms for sequence and structure optimization. Students are introduced to modern structure‑prediction tools and major software platforms such as AlphaFold, Rosetta, FoldX, and AI‑based methods. The course further examines strategies for engineering protein–ligand and protein–protein interfaces. Real‑world case studies highlight applications in enzyme redesign, therapeutic protein engineering, and de novo protein creation are also discussed.

Upon completion the students will be able to:

• Explain protein structural organization from primary to quaternary level and describe the physicochemical forces that drive folding and stability.
• Interpret how structural and energetic principles form the conceptual basis for computational protein design
• Describe the components and assumptions behind physics‑based and statistical scoring functions used to evaluate protein conformations.
• Assess how different energy terms influence structural prediction accuracy and design decisions.
• Compare core computational optimization strategies such as DEE, Monte Carlo, and evolutionary algorithms.
• Select appropriate algorithmic approaches for exploring sequence space efficiently while maintaining biochemical realism.
• Summarize classical and AI‑driven approaches for predicting protein structure from sequence.
• Evaluate how folding simulations can validate, refine, or challenge computational design hypotheses.
• Identify the capabilities, workflows, and differences between platforms such as Rosetta, FoldX, PyRosetta, and AlphaFold.
• Apply common software‑specific workflows for backbone design, stability optimization, and interface engineering
• Explain how machine learning models—such as generative sequence models and structure‑embedding neural networks—are used in protein design.
• Assess the role of data‑driven approaches in accelerating discovery of novel or functional protein architectures
• Describe computational strategies for engineering binding specificity, affinity, and selectivity in protein interfaces.
• Apply concepts of interface energetics to real examples such as inhibitor design, antibody engineering, or creation of synthetic interaction networks.
• Analyze real‑world examples of enzyme redesign, de novo protein creation, and therapeutic protein engineering.
• Translate computational predictions into hypotheses suitable for experimental validation in biotechnology or therapeutic development.

Module Syllabus

• Foundations of Protein Structure and Energetics
• Energy Functions and Scoring Models in Protein Design
• Algorithms for Protein Sequence Optimization
• Structure Prediction and Folding Simulations
• Rosetta and Other Protein Design Software Suites
• Machine Learning and AI in Protein Engineering
• Computational Design of Protein–Ligand and Protein–Protein Interfaces
• Case Studies and Applications in Therapeutics and Biotechnology

Suggested Bibliography

• Relevant literature per lecture, including scientific publications and reviews from international journals, which is available in the course e-class.
• Principles and techniques of Biochemistry and Molecular Biology, 7th edition, Edited by Keith Wilson and John Walker, electronic source