EUVIMEDEUVIMED entdecken
Crossref

Neural Networks for Learning Macroscopic Chemotactic Sensitivity from Microscopic Models

Radek Erban

SIAM Journal on Life Sciences · 31.03.2026 · 1 (1) · 121-141

DOI: 10.1137/25m1799064 ↗

Kurz gefasst

Fachartikel in der Zeitschrift SIAM Journal on Life Sciences, online erschienen am 30.03.2026. Verfasst von Radek Erban. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 52 Literaturangaben. Gefördert von: Engineering and Physical Sciences Research Council.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The macroscopic (population-level) dynamics of chemotactic cell movement—arising from underlying microscopic (individual-based) models—are often described by parabolic partial differential equations (PDEs) governing the spatio-temporal evolution of cell concentrations. In certain cases, these macroscopic PDEs can be analytically derived from microscopic models, thereby elucidating the dependence of PDE coefficients on the parameters of the underlying individual-based dynamics. However, such analytical derivations are not always feasible, particularly for more complex or nonlinear microscopic models. In these instances, neural networks offer a promising alternative for estimating the coefficients of macroscopic PDEs directly from data generated by microscopic simulations. In this work, three microscopic models of chemotaxis are investigated. The macroscopic chemotaxis sensitivity is estimated using neural networks, thereby bridging the gap between individual-level behaviors and population-level descriptions. The results are compared with macroscopic PDEs, which can be derived for each model in certain parameter regimes. Relevance to Life Sciences. Many unicellular organisms exhibit chemotaxis, moving in response to extracellular chemical cues, such as nutrients or toxins. They detect these cues via membrane-bound receptors, and binding to these receptors triggers intracellular signaling cascades that bias their movement. Stochastic, individual-based modeling can be used to capture the fine-grained dynamics at the single-cell level. In this work, we employ detailed, individual-based modeling of chemotaxis to infer population-level properties that characterize collections of cells. Mathematical Content. At the individual level, the mathematical frameworks for chemotaxis include (i) position-jump processes (discretized stochastic differential equations), (ii) velocity-jump processes, and (iii) velocity-jump processes with internal dynamics. The first describes a cell by a stochastic process for its position, the second describes a cell by both its position and its velocity, while the third also introduces variables that describe intracellular processes. Population-level behavior is inferred either by deriving macroscopic PDEs or by training feedforward neural networks on data obtained from simulations of microscopic, individual-based models.

Förderung

  • Engineering and Physical Sciences Research Council

Zitieren

APA

Erban, R. (2026). Neural Networks for Learning Macroscopic Chemotactic Sensitivity from Microscopic Models. SIAM Journal on Life Sciences, 1(1), 121-141. https://doi.org/10.1137/25m1799064

Vancouver

Erban R. Neural Networks for Learning Macroscopic Chemotactic Sensitivity from Microscopic Models. SIAM Journal on Life Sciences. 2026;1(1):121-141. doi:10.1137/25m1799064

BibTeX
@article{Erban2026,
  author = {Erban, Radek},
  title = {Neural Networks for Learning Macroscopic Chemotactic Sensitivity from Microscopic Models},
  journal = {SIAM Journal on Life Sciences},
  year = {2026},
  volume = {1},
  number = {1},
  pages = {121--141},
  doi = {10.1137/25m1799064}
}
Kostenlos starten →