About
Structural Bioinformatics Course 2023
Hey everyone! This is a course about structural bioinformatics and tools you can use it to visualize, analyze and design proteins and other structures. It is taught in the biochemistry undergraduate course at Heidelberg University, but the material is open to be used by everyone. By the end of this course, you will be able to read discuss about machine learning approaches applied to structural bioinformatics, implement your ideas and evaluate the results. You will also be able to use the tools and methods we will discuss to analyze and design proteins and other structures.
This course (Structural Bioinformatics, SB) is designed to be taken in a semester-long format, but can be completed in a shorter time frame. The course is divided into 11 modules with accompanying videos and exercises. The idea is that the components in this course are complementary to each other; in some lessons you will learn more about the underlying machine learning concepts, in others you will see how these concepts are applied to biological problems. On top of that, you will get hands-on experience with some of these algorithms in the case studies and exercises provided.
The course content of the different weeks together with the corresponding case studies we will look at are listed below:
| Week | Course Content | Case Studies |
|---|---|---|
| 1 | Introduction | PyMol, Python, Jupyter Notebooks |
| 2 | ML Basics | Google Colab, PyTorch |
| 3 | ML Architectures | AlexNet, transformers |
| 4 | Language, Evolution and Bioinformatic | ESM |
| 5 | Geometric Deep Learning (GNNs) | PyTorch |
| 6 | Protein structure prediction | AlphaFold2, EMSFold |
| 7 | Generative Modelling | VAEs, Diffusion Models |
| 8 | Protein Design | RFDiffusion, ProteinMPNN, AlphaFold2 |
| 9 | Simulations | GROMACS, Allegro |
| 10 | Drug Design (Docking and Generative) | AutoDock, DiffDock, DiffSBDD |
| 11 | Further topics and conclusion |
I hope you enjoy the course! In case you have any questions or feedback (or find any mistakes in the material), please feel free to contact me.
Whereas the main reference for the classical biochemistry courses is the Lehninger, the content of this SB course is evolving so rapidly that there is no canoical textbook yet. Therefore, many of the resources we will work with will be online tutorials, blog posts as well as original research papers. In addition, this course heavily builds on the material from the following resources:
Publications
- Introduction to Structural Bioinformatics
- A Semester-Long Learning Path Teaching Computational Skills via Molecular Graphics in PyMOL
- Development of a Broadly Accessible, Computationally Guided Biochemistry Course-Based Undergraduate Research Experience, developed by members of the Rosetta Commons project
- Computer-Aided Drug Design for Undergraduates
- Foldit Standalone: a video game-derived protein structure manipulation interface using Rosetta
- Increasing Computational Protein Design Literacy through Cohort-Based Learning for Undergraduate Students, again developed by members of the Rosetta Commons project
- PyRosetta Jupyter Notebooks Teach Biomolecular Structure Prediction and Design
- Teaching structural bioinformatics at the undergraduate level
- Developing and Implementing Cloud-Based Tutorials That Combine Bioinformatics Software, Interactive Coding, and Visualization Exercises for Distance Learning on Structural Bioinformatics
- Introduction to artificial intelligence and deep learning using interactive electronic programming notebooks
- TeachOpenCADD: a teaching platform for computer-aided drug design using open source packages and data
Practicals/Notebooks
- Structural Bioinformatics Practical, Alexey Morgunov
- Cloud-based Tutorials on Structural Bioinformatics
- PyRosetta Notebooks
- Introduction to AI via Cheminformatics