Speakers
Description
This workshop introduces Catalyst.jl, a feature-rich Julia library for modeling chemical reaction networks (CRNs) arising in epidemiology, systems biology, ecology, and beyond. In Catalyst, symbolic reaction systems can be constructed using chemical reaction notation or generated programmatically, and can be coupled to non-reaction ODEs, algebraic equations, and events. Catalyst models automatically convert to symbolic ModelingToolkit.jl representations of reaction rate equations (ODEs), chemical Langevin equations (SDEs), stochastic chemical kinetics (jump processes), or hybrid models. These are then compiled to optimized Julia functions, enabling access to a broad variety of solvers and integration with higher-order tools (such as for sensitivity analysis, parameter estimation, and bifurcation analysis).
Catalyst's symbolic representation enables a variety of CRN analysis, including detection and elimination of conservation laws, calculation of network properties (such as deficiency), automatic calculation of dependency graphs for Gillespie methods, and algebraic simplifications that significantly speed up computations.
The first half of this workshop will focus on model creation and simulation. The second half will demonstrate tools for model analysis and inverse problems, including symbolic computation of network properties (e.g., deficiency indices), steady-state analysis via homotopy continuation, bifurcation analysis, and parameter estimation via PEtab.jl.
