#39 · Systems biology

In silico mechanistic models optimize chemotherapy, ESA and cancer vaccine regimens

The common thread is mechanistic, model-based optimization of therapeutic regimens in silico. The chemotherapy and radiotherapy study shows that the mathematical form of a treatment model can bias model-optimized predictions, and it argues for identifiability and uncertainty analysis over information criteria when selecting models. The other two preprints apply this paradigm to specific therapies: a hybrid AI-MIDD/QSP digital-twin of erythropoiesis, encoded in SBML, simulates ESA therapy variants and optimizes combinations, while a multiscale PBPK model of tumor vaccination links lymphatic transport, antigen-presenting cell activation and immune priming to optimize vaccine timing. What the group adds is attention to structural model choice and identifiability as prerequisites for trusting optimized dosing and scheduling. Items are heterogeneous in scope, so the shared element is methodological rather than a common biological system.