Skip to content

Latest commit

 

History

History
95 lines (59 loc) · 9.99 KB

README.md

File metadata and controls

95 lines (59 loc) · 9.99 KB

Introduction

This is a list of resources (courses, tutorials, etc.) in the field of Computational and Structural Biology. This is for those with no background in biology but who want to work in this field.

I highly expect you to contribute to this project if you have any ideas on related topics. One person can go fast, but a group of people can go further.

Structural Biology

The first step is to learn the very fundamental concepts in biology, like what is an amino acid? What are proteins made of? You don't have to memorize every detail, but it is essential to be familiarized with those conceptual words. After this stage, you should understand the basic structure of proteins, e.g., the four levels of structure.

Computational Biology

Alphafold 2

'Morden' computational methods for biology are dominated mainly by neural networks, the most famous is Alphafold2.

Alphafold2 achieves an atomic level of accuracy in the protein structure prediction task. It gives a solution to a problem that has not been answered for 50 years. The model capacity is gigantic, and the design is sophisticated. To understand the paper, I list several tutorials that describe the mechanisms of Alphafold2. I think that many operations in Alphafold2 are no biologically or physically meaningful; instead, they just try to find a method to scale up the model capacity effectively and thus achieve better performance. But it does not mean I don't like Alphafold2. The uses of the structure module and the representation of the atom's 3D coordinate are as brilliant as hard to understand.

Practices

"What I cannot create, I do not understand." -Richard Feynman.

As I have stated, most current methods focus on developing neural networks for bio problems. If you have a fair understanding of deep learning, you can find a paper to read and start your project. If you are a beginner in deep learning (DL), there is a long way to go. Many good materials on deep learning are publicly available; please look them up. Either way, I recommend you reproduce a SOTA method in the field of computational biology. If you don't know what paper to reproduce, I think Alphafold2 is a great starting point! That will ground you up for your project.

I also list several beginner-level excises below.

Cutting-edge research