Pulling your model up by its bootstraps: evolving your way to QSP

QSP models present oncology modelers with a chicken-and-egg conundrum: how to build a QSP model for a novel target, when cancer is a set of molecularly distinct diseases, posing significant challenge to building or extending a platform-style model that spans multiple targets? In addition, the desire to quickly progress drug programs to the clinic in this area of high unmet medical need leaves little time or resource for building large models from scratch. We’ll review case studies illustrating how to progressively enhance model detail as drug projects move from in vitro (lead generation) to in vivo (lead optimization) to nomination and beyond, aiming towards a QSP model that separates the properties of compound and system/target.