Title: Sparse model identification and prediction of microglial cells during ischemic stroke

Abstract: Dynamics between key neuroinflammatory components, detrimental M1 and beneficial M2 microglial cells, are not fully understood post-ischemic stroke. To discover, model, and predict these dynamics, we use a method based on sparse identification of nonlinear dynamics (SINDy). The resulting data-driven dynamical system involves constant and linear terms but does not include nonlinear interactions between cells. Results show M2 microglial cell dominance of four days. Forward predictions capture potential long-term dynamics of microglial cells and suggest a persistent inflammatory response.