Behavioural biology

Modelling biological organisms is a challenge both for physics and for biology, due to the difficulty of finding a good balance between the level of abstraction with which individuals are described and the explanatory power of the model. The artificial intelligence (AI) – or artificial agency – paradigm opens up new possibilities to enrich the description of the individual organisms and to model processes as complex as evolution from a radically different perspective. We use Projective Simulation (PS) to model biological entities as artificial learning agents that interact with each other and with the environment in a reinforcement learning framework. This new approach allows us to study fascinating phenomena such as the collective motion [1, 2] of foraging animals, e.g. swarming locusts, or the collective defence of honeybee colonies [3]. Evolutionary pressures arising from the scarceness of food or the attack of a predator can be encoded in the reward function, so that the iterative interaction of the ensemble of agents with the environment can be interpreted as an evolutionary process.

Collective motion in foraging scenarios

locusts

The well-known phenomenon of collective motion, observed for example in swarming locusts, schooling fish and many more animals, has been extensively studied for decades. Modern tracking technologies have helped to improve our understanding of how individual interactions can give rise to collective behaviour [*]. In our group, we have applied PS to model behavioural experiments such as marching locusts in a ring-shaped arena [1], or more general scenarios, in which ensembles of agents forage in resourceful or resource-scarce environments [2].


Collective defence of honeybee colonies

bees

To sting or not to sting, that is the question. In an exciting collaboration with biologists from the University of Konstanz, we have studied the individual responses of honeybees to the so-called alarm pheromone, a compound that is released when the bee stings and alerts the colony about possible intruders [3]. We model the colony as an ensemble of PS agents that undergoes repeated encounters with predators and whose success, as a colony, depends on the remaining live bees, considering that a bee dies both when stinging and when it’s killed by the predator. The collective defence is built up as a sequence of individual decisions in which each agent decides whether to sting or not.


Quantum Biology

The question, to which extent quantum coherence is exploited in biological systems, e.g., in order to enhance their efficiency, has attracted much attention in the recent years. Examples of interest are in the process of light harvesting in photosynthesis, or the radical pair mechanism as a model for avian magneto-reception. In our research, we are exploring the general conditions for the existence of quantum coherence and entanglement in biological systems, as well as their possible role for biological function.

  • We study the models of non-equilibrium quantum systems that are capable of generating and sustaining entanglement in a noisy environment. We are interested in such models both from the perspective of building a noise-resilient quantum computer that could operate under room-temperature conditions, and from the perspective of identifying entanglement-sustaining molecular-scale mechanisms in biological systems.    
  • We study the radial pair mechanism in spin chemistry, which is believed to play a central role in avian magneto reception. We investigate the effect of multiple encounters and other sources of de-coherence in the radical pair mechanism, using both a stochastic collision model and a master-equation approach.
  • We explore the use of molecular photo-switches to optically control the radical pair reactions and to test current theory used to describe spin-recombination reactions.   

 


  • [1] K. Ried, T. Müller, and H. J. Briegel, Modelling collective motion based on the principle of agency: general framework and the case of marching locustsPLoS ONE 14(2), e0212044 (2019) [arXiv:1712.01334].
  • [2] A. López-Incera, K. Ried, T. Müller, and H. J. Briegel, Development of swarming behavior in artificial learning agents that adapt to different foraging environmentsPLoS ONE 15(12), e0243628 (2020) [arXiv:2004.00552].
  • [3] A. López-Incera, M. Nouvian, K. Ried, T. Müller, and H. J. Briegel, Honeybee communication during collective defence is shaped by predationBMC biology 19, 106 (2021)  [arXiv:2010.07326].
  • [*](external) A. I. Dell et al., Automated imaging-based tracking and its application to ecologyTrends in Ecology and Evolution 29, 417-428 (2014).
  • [4] H. J. Briegel and S. Popescu, A perspective on possible manifestations of entanglement in biological systems, in Quantum Effects in Biology, eds. M. Mohseni, Y. Omar, G. S. Engel, M. B. Plenio, Cambridge University Press (2014).
  • [5] J. Clausen, G. G. Guerreschi, M. Tiersch, and H. J. Briegel, Multiple re-encounter approach to radical pair reactions and the role of nonlinear master equationsJ. Chem. Phys. 141, 054107 (2014) [arXiv:1310.6194].
  • [6] G. G. Guerreschi, M. Tiersch, U. Steiner, and H. J. Briegel, Optical switching of radical pair conformation enhances magnetic sensitivityChem. Phys. Lett. 572, 106 (2013) [arXiv:1206.1280].
  • [7] M. Tiersch, S. Popescu, and H. J. Briegel, A critical view on transport and entanglement in models of photosynthesisPhil. Trans. R. Soc. A 370, 3771 (2012)[arXiv:1104.3883].
  • [8] J. M. Cai, S. Popescu, and H. J. Briegel, Dynamic entanglement in oscillating molecules and potential biological implicationsPhys. Rev. E 82, 021921 (2010)[arXiv:0809.4906].
Nach oben scrollen