Dresden 2026 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
BP: Fachverband Biologische Physik
BP 14: Poster Session II
BP 14.7: Poster
Dienstag, 10. März 2026, 18:00–21:00, P2
Modeling host-pathogen interactions via stochastic simulations and neural network-driven Bayesian inference — •Soham Mukhopadhyay1, Jonathan Pollock2, David Voehringer2, and Vasily Zaburdaev1 — 1MPZPM, Erlangen, Germany — 2Department of Infection Biology, University Hospital Erlangen, Friedrich-Alexander University Erlangen, Germany
Helminth infections affect a large proportion of the world’s population and cause significant morbidity. There are no vaccines against helminths, and the mechanisms of anti-helminth immune responses are often not well-understood. Taking the murine hookworm N. brasiliensis as our experimental system, we develop a mechanistic model that describes the parasite load in different host organs — which the parasite migrates through over the course of its lifecycle — as a function of time. We abstract infection progression as a state-transition process and simulate it via kinetic Monte Carlo, thereby linking the infective dose of larvae to the number of eggs shed to the environment by adult worms from the host intestines, which can then be compared to experimental data. To infer model parameters — the various transition rates — from experimental data, we employ emerging techniques from the domain of neural network-driven Bayesian inference to infer the posterior probability distribution of parameters conditioned on data. Using this model and inference framework, we plan to compare the population dynamics for different immune perturbations.
Keywords: Simulation-based inference; Population dynamics; Neural Bayesian Inference; Kinetic Monte Carlo