Description du projet
Un regard plus attentif sur les cellules amacrines dans les calculs visuels naturels
L’une des problématiques les plus complexes en neurosciences visuelles est l’énorme diversité des cellules amacrines (CA), une classe d’interneurones inhibiteurs présents dans la rétine des vertébrés. Malgré leur importance, les CA restent l’une des classes de neurones de la rétine les moins bien comprises. Afin de mieux appréhender leur rôle dans les calculs neuronaux, divers modèles ont été développés à différents niveaux de réalisme. Le projet NextMechMod, financé par l’UE, s’appuiera sur les progrès de l’apprentissage automatique pour créer une nouvelle génération de modèles mécanistes hybrides. Ceux-ci s’appuieront sur l’intégration de différents niveaux de réalisme dans la modélisation. Par ailleurs, le projet concevra un ensemble d’outils qui permettront d’étudier systématiquement le rôle des CA rétiniennes au cours des calculs visuels naturels.
Objectif
Ever since the work Hodgkin and Huxley, models of neurons have been essential for our understanding of neural computations. Such models have been developed at diverse levels of realism, from linear-nonlinear cascade or black-box models to detailed compartmental models. While these approaches are commonly viewed as incompatible, they have attractive strengths from an epistemic point of view. In this project, I propose to develop a new generation of hybrid mechanistic models that reconcile these levels of modelling: they will consist of a compartmental model for the neuron of interest with inputs approximated by black-box models. I will leverage the power of these hybrid models to tackle one of the most challenging questions in visual neuroscience: the staggering diversity of amacrine cells, a major class of inhibitory interneurons in the vertebrate retina. Despite their diversity, they are the least understood class of neurons in the retina, in stark contrast to the remaining circuitry. While in mouse more than 60 types of ACs have been identified by single cell transcriptomics, only a handful has been studied at depth. I will build on the latest advances in machine learning to develop a framework for efficiently inferring the parameters of a hybrid mechanistic model. To constrain the model parameters, we will acquire two-photon calcium and voltage imaging data during natural stimulation. Further, we will extend our framework to incorporate transcriptomic information about gene expression collected via patch-seq into the inference procedure, allowing us to map the amacrine cells to genetically defined types. Thus, in this project, I propose to develop a toolset to systematically uncover the role of retinal amacrine cells during natural visual computations, and link it to its mechanistic basis, providing a path forward to solving one of the key remaining mysteries of visual neuroscience.
Champ scientifique
Programme(s)
- HORIZON.1.1 - European Research Council (ERC) Main Programme
Thème(s)
Régime de financement
HORIZON-AG - HORIZON Action Grant Budget-BasedInstitution d’accueil
72074 Tuebingen
Allemagne