Michael Pun is a PhD candidate in the department of physics at the University of Washington. In 2017, he obtained his bachelor’s degree in physics from Bowdoin College, Maine. In 2018, he was awarded a doctoral grant from the Max Planck Institute of Dynamics and Self-Organization, Germany, and spent a year there. His interests lie at the intersection of physics, structural biology, and machine learning. Specifically, his research focuses on incorporating physical symmetries into machine-learning algorithms to build more efficient, robust, and interpretable data-driven models of protein structures. He recently developed a novel model of amino-acid preferences in proteins that can predict mutational effects on protein stability and binding.