CRCNS: Using Perturbations and Modeling to Study Connectivity for Decision-Making - Much work over the past decades has identified brain areas that are necessary for decision-making and neural activity patterns that correlate with decision-making computations. These findings have inspired computational models that propose greatly differing computations (e.g., competition, amplification, and stabilization) for how the same cortical circuits contribute to decision-making, and thus distinct underlying connectivity mechanisms. While much experimental work has tested and refined these models, from neural activity measurements alone, it often remains unclear if a brain area generates its decision-related activity patterns, and how it might generate them, or merely inherits them from another brain area. Ultimately, a direct measure of the connectivity motifs, and the computations they perform, is needed to identify what and how cortical circuits contribute to decision-making. Here, we propose to use cellular-resolution optogenetics to measure a circuit's effective connectivity by perturbing small groups of neurons while measuring the effect on neural activity in surrounding neurons. In the first aim, we will measure the effective connectivity between excitatory and inhibitory neurons with identified activity patterns during a decision task in virtual reality. In the second aim, we will develop and test multiple classes of computational models in conjunction with the experimental results. We aim to discover the connectivity motifs that exist in cortex and reveal the decision-related computations that are performed by these motifs. We hypothesize that recurrent connections amplify choice selectivity through competition motifs mediated by inhibitory neurons and create choice-informative sequences of neural activity through inhibitory control of activity timing. We thus predict that inhibitory neurons play a critical role in these computations by having sharp choice selectivity and precise connectivity profiles. RELEVANCE (See instructions): The proposed work will reveal fundamental mechanisms of neural function for decision-making. The new methods, models, and concepts will provide a foundation to understand how decision-making is disrupted in neurological disorders and diseases.
