Speakers
Description
Biochemical reaction networks (BCRNs) form the backbone of systems biology. Effectively modeling complex biological systems requires a delicate balance between complexity and accuracy, motivating the need for rigorous, principled BCRN reduction. This minisymposium addresses the demand for systematic and algorithmic, model-independent reduction methods. Featured topics include lumping, quasi-steady-state approximations (QSSA), and Tikhonov-Fenichel reductions, with a specific focus on the parametrisation method for systems exhibiting multiple timescales.
Participants will present algorithms to non-dimensionalise and convert large reaction systems into forms amenable to formal mathematical analysis. The session explores how lumping aggregates subsets of state variables into new ‘lumped’ variables to enable exact dimensionality reduction while retaining dynamical fidelity. Furthermore, the minisymposium highlights how algebraic singular perturbation theory and coordinate-independent geometric singular perturbation theory (GSPT) provide a robust framework for reducing multiscale BCRNs without relying on rigid, heuristic timescale separations. Finally, the minisymposium examines the total quasi-steady-state approximation (tQSSA) for deterministic and stochastic networks, alongside applications in pharmacokinetics. By replacing intuition with algorithmic rigor, these methods yield mathematically sound reductions that enhance both analytical insight and computational tractability.
