Bayesian framework is ideal for this type of inference as it allows us to combine population and personal effects in a principled way and make predictions for both groups and individuals. The inferences are further improved when we introduce mechanistically inspired components into the modeling framework.

In this talk, I will provide an overview of these types of models using three real-world examples:

1) Pharmacokinetic model suitable for an early stage trial in small patient populations, particularly in rare diseases

2) Semi-mechanistic joint survival model for targetted therapies in solid tumors