We aim to understand how computation in neural circuits gives rise to flexible, complex behaviour
The brain is remarkable in its ability to produce a rich array of inbuilt and learned behaviours, which are deployed flexibly to meet individuals’ needs and environmental demands. Behaviour emerges from distributed computations in specialised neural circuits across brain regions, with the same regions often contributing to multiple behaviours. How networks of neurons give rise to this adaptive and complex repertoire of function is a fundamental scientific question of our era.
We seek to identify a set of elemental neural computations that underlie different behaviours and determine how they are implemented at the level of circuits, cells and synapses. Our long-term goal is to use this knowledge to build a coherent theoretical framework that combines these neural computations and explains how complex behavioural processes relate to neural circuit mechanisms. To provide this deep understanding about how activity in neural circuits encodes the fundamental processes underlying behaviour, our research focuses around the following challenging questions:
Because these questions cannot be answered by any one laboratory, we are developing new models of team science, as well as developing exciting new technologies, such as high density Neuropixels probes, which permit recordings from thousands of neurons across multiple brain regions during behaviour. These and other emerging technologies will generate entirely new types of datasets at the spatiotemporal scale relevant for distributed computations underlying behaviour, and therefore help inform and constrain the theories that explain how brain-wide activity leads to behavioural decisions. Through multi-disciplinary team research, and in close collaboration with the Gatsby Computational Neuroscience Unit (GCNU), scientists at the SWC are generating new theories that relate computational algorithms to identified neural circuits to explain different aspects of cognition and behaviour.
Our approach requires investigation at multiple scales, revealing computations performed by synapses, cells, circuits and brain regions. We use methods to identify, label, manipulate and record identified neurons in multiple brain regions during behaviour. This is achieved through the application of state of the art methods including two-photon and wide-field calcium imaging, fiber photometry, electrode array and whole-cell recordings, genetics, anatomy and connectomics, optogenetics and pharmacogenetics, quantitative behavioural methods and real-time brain-computer interfaces (closed-loop, virtual reality).