Deep Learning for Genome Design

Rapid engineering of microorganisms (e.g., E. coli, S. cerevisiae) is currently hindered by limited integration of manufacturing constraints, goal optimization into the design process, reducing the yield and the productivity of several genome engineering workflows. This thesis proposal wants to formalize the genome design as a machine learning problem aiming at finding the Pareto surfaces (i.e., efficient and robust trade-offs) between goal optimization, manufacturing constraints and design conditions. The thesis will address the design of a computational frameworks to rational engineering of genome microorganisms. Skill: Python. The thesis will be conducted in collaboration with European Brain Research Institute Rita Levi-Montalcini - EBRI, Rome.