Work in our lab is motivated by a simple question: what makes animals choose one option over another? We aim to answer this question both at the level of genes and at the level of neural circuits. At the level of neural circuits we aim to link cell-type specific computations with behavioral choices that animals make. By employing computational models we hope to uncover hidden decision variables and by optically tagging specific cell-types we aim to connect those variables to circuit level computations in behaving rodents.
In order to discover the molecular basis of choices we develop automated and trial based assays in fruit flies that mimic natural foraging decisions. The power of fruit fly genetics allows us to carry out large-scale genetic screens to identify single molecules that play a key role in those decisions.

Another focus of our research is to determine whether humans use the same foraging strategies as flies and rodents. Our goal is to identify if conserved genes and neural circuits underlie value based choices both in animals and humans.

Our long-term goal is to use ecology inspired quantitative behavioral models to guide and sharpen scientific questions.

At Aarhus University we took advantage of the diverse scientific culture and started interdisciplinary collaborations with optics groups to develop novel light stimulation and read-out tools for neuroscience.

The methods used in our lab include molecular genetics, psychophysics, behavioral electrophysiology, optogenetics and computational modeling.

We have set up a trial based single fly reward foraging assay to study how animals' choices are affected by reward history. In our behavioral set up animals are freely walking forth and back in a linear arena. At the end of the arena (yellow rectangle) animals experience reward delivered by optogenetic activation of sugar receptors on the labellum of the fly in a probabilistic fashion. This allows us to estimate how reward history affects a flys' decision to enter the rewarded area (yellow rectangle).

Similar to flies we have established a dynamic foraging task in mice where water rewards are given to animals either in a left or a right port in a probabilistic fashion. Over the course of the behavioral session reward probabilities change and we monitor how animals adjust their choices to changing probabilities. In further experiments we use electrophysiological recordings from neurons in mouse prefrontal cortex to reveal how cells in this area represent value of upcoming options.

In collaboration with Jacob Sherson we want to design a human foraging computer game.
We want to explore whether humans and fruit flies use similar strategies to forage in an uncertain environment. We use virtual maze navigation to study foraging strategies for probabilistic rewards in humans.

Current state of the art deep brain optogenetic stimulation is only possible by using optical fibers. However in all optogenetic studies light scatters uniformly from the fibers' end, activating synchronously all light sensitive neurons. This poses a challenge for interpretation of all optogenetic behavioral manipulations. Here in collaboration with optics groups at Aarhus University we aim to develop a patterned light stimulation tool to address individual neurons with the natural neural ensemble activity in behaving animals. This tool should make it possible to “replay” the pattern of neural activity to animals and probe the neural code for perception, decisions and high-level cognition in animals.