Multiscale approximations for stochastic reaction networks

Project description

Stochastic (chemical) reaction networks are models for biophysical systems where each reaction describes a conversion of one or more chemical species to one or more other chemical species. Examples include Michaelis--Menten enzyme kinetic reactions, Togashi--Kaneko autocatalytic reactions, transciption/translation reactions etc. At the ecological scale, susceptible-infected-removed (SIR) epidemic models, predator-prey models could also be considered examples of stochastic reaction networks. The goal of this project is to provide rigorous probabilistic justification to various multiscale approximations such as the quasi-steady-state approximations (QSSAs) derived from deterministic ordinary/partial differential equations. We will also devise efficient simulation algorithms and statistical inference methodologies based on those multiscale approximations.