Augustin co-developed CellBox, a machine learning–based algorithm that models cell dynamics to predict responses to a range of perturbations—a capability that could aid the discovery of combination cancer drug therapies. However, these models do not fully capture protein-level information, and most cancer drugs target certain proteins. As a BroadIgnite awardee, he aims to generate large-scale single-cell proteomics data and leverage machine learning algorithms (like CellBox) to better understand the biological mechanisms driving cancer, and identify new treatments that can mitigate drug resistance.