Abstract
Model-based decision making requires representation of predicted states that are updated by action-dependent state transition models. To investigate their neural implementation, mice were trained to do a virtual navigation task and neural activity was recorded in the posterior parietal cortex (PPC) with the genetically encoded calcium indicator GCaMP6f and 2-photon microscopy.
A mouse was head restrained and maneuvered a spherical treadmill. 12 speakers around the treadmill provided an auditory virtual environment. When the mouse reached a virtual sound source and licked a spout, it got a water reward. The task had two conditions: continuous condition in which the guiding sound was presented continuously and intermittent condition in which the sound was presented intermittently.
We recorded activities of up to 600 neurons simultaneously in layers 2, 3 and 5 of PPC. From population activities, we decoded the distance to sound source; the predicted distances had no significant differences between continuous and intermittent conditions. The predictions were thus preserved irrespective of auditory inputs, suggesting the important role of PPC in action-dependent state prediction.
