Cryo-electron microscopy (cryo-EM) is a Nobel-prized technology that aims at characterizing the 3D structure of proteins at the atomic level. Recent advances in iterative reconstruction frameworks have opened the door to the introduction of learning into the reconstruction process. In that context, the goal of this project is twofold. First, the student will create a large training dataset of cryo-EM data from publicly available databases. Then, the student will train a dictionary learning algorithm with this dataset to learn the 3D structure of the proteins of interest. Good Matlab skills are a requisite for this project.