Neurons are morphologically very complex cells with large arborisations that can extend for several millimetres. Communication between neurons relies on action potential driven release of neurotransmitter at specialised sites called synapses. For many years, the analysis of neuronal circuits used simplified morphological models describing synapses as a single contact point between two neurons. Recent studies, have shown the importance of neuronal compartments and their unique biophysical properties in conveying information in the nervous system. More morphologically accurate models have been developed, but these are often limited to studying signal propagation within a single neuron. This project will use experimental data to build a biophysically accurate model of synaptic connections between neurons that will take in account morphological features and their impact on signal transmission. Of particular interest, a phenomenon called analog-digital signalling has been recently demonstrated to have a morphology-dependent effect on synaptic transmission.
Electrophysiological and morphological data from pairs of connected neurons will be used to produce a functional minimal neural network that can be modelled with neurons with multiple compartments. The detailed model allows for studying ion channels and synaptic components which can be measured by experiments. In this project, the experiments and models are working hand by hand. In silico experiments will be aimed at dissecting the crucial biophysical and morphological properties of neurons that allow flexible, yet reliable, communication. In vitro and In vivo experiments will study the impacts of these crucial components of neurons on the dynamics of single neurons and connected neurons.
With the recent development of computational techniques in machine learning and artificial intelligence, it has been suggested that signal propagation within a single neuron with a morphological structure of multiple dendrites can be used to solve some tasks, for example, credit-assignment. However, one knows that computations are mainly done and carried out with a network of neurons in the brain. This project is the first step towards to understanding neural computations at a circuitry level but with morphological significance of single neurons.
Techniques that will be undertaken during the project
-Neuronal modelling with morphological structure -Neural network modelling with synaptic plasticity -Patch-clamp and other electrophysiological techniques and their quantification -Fluorescence microscopy