The Darshan Lab: Theoretical and Computational Neuroscience Lab
Research
In the lab we embark on an exciting journey to unravel the mysteries of the brain's computational capabilities. Using computer simulations and mathematical tools, our research delves into the intricacies of neural circuits, aiming to understand how cognitive abilities emerge from collective neural activity and evolve during the learning process. We collaborate closely with experimentalists, employing diverse theoretical and computational tools to shed light on design principles and mechanisms underlying brain computations. From memory and decision-making processes to sensory perceptions and motor outputs, our work spans a wide range of neuroscience domains, seeking to build comprehensive models of neurons and neuronal networks to elucidate the neural basis of cognition.
Current Research Directions
Multi-regional computations in the brain
Recent advances allow to record neurons from many places in the brain at the cellular level. We are collaborating with multiple labs to build neural circuit models of brain-wide functions. Based on these collaborations, we look for organizational principles of multi-regional computations in the brain.
Identifying plasticity rules from neural representations and behavior
A dominate dogma in neuroscience is that learning results from changes in synaptic connections. While various synaptic plasticity rules have been proposed to explain these changes, it remains a challenge to connect them to the neural representations and behaviors observed in learning animals. We leverage cutting-edge techniques from the field of Artificial Intelligence (AI) to develop novel methods for discovering plasticity rules from neural representation changes that occur during the learning of new behaviors.
​
Build theoretical frameworks to understand neural circuit computations through perturbations
New technologies allow to study brain functions and their underlying network structure by perturbing the dynamical state of neural networks in vivo in various conditions. However, due to the non-linear nature of neural networks, it is notoriously difficult to interpret these experiments without mathematical models and theories. We develop such theories and test them together with experimental collaborators.​
​
In addition, we look for any novel and exciting ideas that exploit mathematical tools to shed light on questions in System Neuroscience
