Research
• Updated
16 December 2021
LNC2

Heike Stein, cognitive neuroscience researcher, EMBO funding winner

Heike Stein has been a post-doctoral student at the Laboratoire de Neurosciences Cognitives et Computationnelles (LNC2) for one year. She has just obtained a European Molecular Biology Organization (EMBO) grant which will allow her to continue her research with Alex Cayco in the "Mathematics of Neural Circuits" team of LNC2 for the next two years.

Heike

Heike Stein started her postdoc in the Group for Neural Theory (GNT) with Alex Cayco-Gajic one year ago, after receiving a PhD from the University of Barcelona, where she worked with Albert Compte for four years. In her bachelor's and master's in psychology and cognitive neuroscience, she developed an increasing interest in statistics and modeling, and in the neural mechanisms that underlie cognition. « Based on that, I decided to work on a project where I could use my background in psychophysics while learning about neural data analysis and computational network models. In my thesis, I investigated how NMDA receptor hypofunction can affect working memory in psychiatric and neurological disease. »

 
Research to understand how brain circuits control cognition and behaviour

Heike Stein is interested in understanding how brain circuits control cognition and behavior. « Different neural circuits are more or less suited for certain types of computations, depending on their architecture and connections with other brain areas. For example, the prefrontal cortex (which I studied during my PhD) is a higher-order structure that receives extensive input from lower, e.g. sensory areas and is characterized by strong recurrent connectivity within the circuit. Thanks to these features, it can sustain ongoing representations of sensory or abstract variables that are needed in working memory tasks. » More recently, heike became interested in circuits with even simpler and more tractable architectures, « which is helpful when formulating hypotheses about their computational roles. The cerebellum for example has a divergent-convergent feed-forward architecture and prominent error-related plasticity within the circuit. Therefore, researchers think that it plays an important role for error-based learning. In my current projects, I am using sophisticated statistical methods to identify signatures of motor learning in cerebellar neural activity. »


EMBO provides support for the life sciences in Europe and beyond

EMBO (European Molecular Biology Organization) is an international organization for life sciences that supports researchers at different career stages with short-term and long-term funding. Heike Stein received an "EMBO Postdoctoral Fellowship" grant, a grant to promote the international mobility of young researchers. « It will pay my salary for two years. Similar to other postdoc fellowships, to be eligible, not more than 2 years should have passed since the PhD and one needs to move to a different country than where the PhD was done. It is favorable to explore new topics rather than to continue working on the same problems as during the PhD. Writing an application costs some time: You will have to write about your background, research interests, and propose a specific research plan to be carried out at the host lab. Yet, writing a proposal usually helps to get a clear idea of what you want to study in your postdoc. »​​​​​​​


Research in the "Mathematics of Neural Circuits" team

This EMBO grant allows the young researcher to work with Alex Cayco-Gajic, junior professo at the ENS and co-directror of the "Mathematics of Neural Circuits" team with Boris Gutkin.  « Alex Cayco started her team two years ago, after a PhD in applied mathematics with Eric Shea-Brown and a postdoc with Angus Silver at UCL, where she discovered her fascination for the cerebellum and for large-scale data analysis. »

In this team, researchers combine a wide range of computational techniques to study cerebellar computations. Different projects range from recurrent neural networks and reinforcement learning models to the use and development of new methods for data analysis. « One of my projects is more on the technical side: Together with a master student, we developed a new dimensionality reduction technique that can uncover slow changes in neural population activity. In my second project, we are working closely with the experimental lab of Michael Graupner at Paris Descartes to study how mice learn regular gaits in novel environments.»


Towards research on cerebellar-cortical interactions during cognition and learning

« Coming from a cognition-centered background, learning about the cerebellum and motor control theories is new and refreshing to me, and in Alex' group I am receiving training that is quite complementary to my background. ». In future research projects, Heike Stein would like to combine her background and this new knowledge to rethink how the cerebellum contributes to functions that are traditionally thought of as purely neocortical. « For example, there is more and more evidence for cerebellar involvement in cognition and learning. At the same time, novel computational techniques make it increasingly accessible to analyze and model the communication of multiple brain areas during cognitive tasks. In the future, it will be an exciting avenue to study cerebello-cortical interactions during cognition and learning. »

 

More information about the Mathematical for neural circuit team (By Boris Gutkin)

"Over the past decades, we have learned a lot about the biology of brain cells - neurons: their electrical and biological properties. In the brain, neurons connect to each other to form complex constellations: interlocking circuits and networks that neuroscientists believe underlie brain function. Much experimental data has been collected on these circuits. In our team, we try to translate biological data into the language of mathematics, by building mathematical models of biology. We then study these models to understand how the properties of individual neurons and the way these cells are interconnected can give rise to the mechanisms that make our brains 'think'. Using these models, we study, for example, how addictions arise from the pharmacological action of drugs, how our memories are formed, and how signals from our bodies are taken into account by our brain systems to shape and mould our motivations."