Modelling stem cell differentiation related processes: A practical overview for biologists

Stem cell differentiation is central to development and regenerative medicine, but the complex underlying processes hinder our ability to control it experimentally. Computational models can help form hypotheses and generate predictions. There are many types of computational models available to aid in understanding stem cell differentiation–related processes, but for an experimental biologist, it might be hard to select the modelling approach that matches their research question, and to know what data would be needed to make use of the model. This review is aimed at experimental biologists to introduce various modelling types, relate them to the types of questions these models can help answer and outline the necessary data to gain new insights. The review discusses mechanistic dynamic models, both ordinary differential equation (ODE) and abstract, multiscale models and data-driven deep learning approaches. Each model class is introduced with what the model represents, the insights it can provide, validation strategies and limitations. With this review, we want to make it easier to incorporate modelling within experimental workflows for stem cell differentiation–related research, to aid experimentation and accelerate discovery.