From structural biology to structural cell biology with the aid of machine learning:
Structural biology has long provided atomic-resolution insights into macromolecular function, but traditional approaches, such as X-ray crystallography and single-particle Cryo-EM, require the purification of molecules, removing them from their native cellular environment. This reductionist approach has been crucial for determining structures, yet it often fails to capture how macromolecular assemblies function dynamically within the cell. Conversely, cell biology techniques provide a broader view of cellular processes but often lack the molecular resolution necessary for mechanistic understanding.
The emergence of Cryo-Electron Microscopy (Cryo-EM) and Cryo-Electron Tomography (Cryo-ET) is now bridging this gap. Cryo-ET, in particular, allows researchers to visualize macromolecular assemblies directly inside cells, at near-atomic resolution, offering unprecedented insight into cellular architecture and macromolecular function in situ. However, analyzing and interpreting the vast amount of data generated by Cryo-ET remains a challenge. Machine Learning methods are now playing a key role in improving structural classification, denoising, and segmentation in Cryo-ET datasets, making them an essential tool for modern structural biologists.
This Course will train the next generation of researchers in these cutting-edge integrative approaches, ensuring that they are equipped with both experimental and computational skills to tackle the challenges of Structural Cell Biology.