The goal of this thesis is to biologically evaluate and refine our existing diffusion-based SELEX simulator.You will work with a ready-to-use generative model and focus on (i) defining meaningful biological metrics, (ii)identifying gaps between simulation and real SELEX outcomes, and (iii) implementing targeted improvements thatincrease realism and usefulness for virtual screening.