| Please join us for a discussion on how we study the shapes of biological molecules, leading up to the new AlphaFold 3 AI model from Google DeepMind and Isomorphic Labs. We will start by talking about the famous protein folding problem and how scientists used to figure out these shapes using older methods like X-ray crystallography and Ramachandran plots. From there, we will quickly review how the first AlphaFold and AlphaFold 2 models were built and why they were such a big deal for science.
Next, we will dive into what makes AlphaFold 3 a major step forward. While older versions mostly focused on proteins, this new model uses an updated design and a diffusion network to predict the 3D shapes of almost all of life’s molecules, including DNA, RNA, and drug molecules. It is much more accurate at showing how all these parts interact with each other.
|
