2025
The collective dynamics of microorganisms are extremely rich, echoing the great diversity of signals that single cells can sense, process and produce. For example, paramecia naturally produce CO2 which locally acidifies the environment and attracts other individuals, but at the same time consume oxygen which has a repulsive effect. These opposing interactions create front dynamics that can lead to the stabilization of aggregates with non-trivial inner geometries (fig 1-a). Another example is that of the microalgae Chlamydomonas reinhardtii, which on the one hand absorb light and on the other hand seek darkness (negative phototaxis), which creates spontaneous aggregations and varied patterns1 (fig. 1-b).
However, at these scales it is not easy to decouple the effects stemming from hydrodynamic forces from those due to the activity of microorganisms in response to interactions. Recent multi-agent simulations allow each agent to be controlled by a simple artificial neural network modeling the internal interaction networks of cells. By bringing large sets of agents into interaction, these simulations reproduced in encouraging ways various aspects of the dynamics of real systems.
This internship thus aims at generating collective dynamics of microorganisms with these multi-agent simulations in order to identify the minimal ingredients necessary for the formation of complex dynamic patterns and to formulate predictions for experimenters. Depending on the progress, it is also envisaged to allow the neural network to evolve and become more complex in order to better reproduce the dynamics observed experimentally.
Bibliography
1. Eisenmann, I. et al., Light-induced phase separation with finite wavelength selection in photophobic micro-algae - ArXiv (2024) https://doi.org/10.48550/arXiv.2401.08394