The OpenScope program opens a large-scale neuroscience experimental setup, the Allen Brain Observatory, to projects proposed by researchers outside the Allen Institute.

We seek to understand to what extent predictive computations are carried out at multiple stages in the cortical hierarchy. To this end, we have designed and carried out a temporal sequence learning experiment. The core of this design involves the presentation of a repeated sequence of three natural movie clips (2 sec duration). By passively presenting this sequence of visual stimuli over 3 days, the mouse’s brain can potentially learn to use events at the end of one movie clip to predict the occurrence of events at the beginning of the next. Before and after temporal sequence training, the same movie clips were presented in random order, intermixed with a 2 second grey screen, in order to characterize neural responses. This design allows comparisons of: 1) neural responses to each movie before and after sequence training, 2) responses for pairs of movies played in the trained vs. untrained order, 3) evolution of responses to movie sequences during training, among others. Recordings were made using two-photon imaging from 8 fields-of-view covering both layer 2/3 and layer 4 in four cortical areas: VISp (V1), VISl (LM), VISam (AM), and VISpm (PM). Experiments were carried out in 13 mice from the Cux2 transgenic line that expresses GCaMP in superficial excitatory neurons.

We asked OpenScope to fill two gaps in the open data ecosystem for mouse visual neurophysiology. First, there was a dearth of recordings from the two deep structures in the brain that receive direct input from the retina. We requested prioritization of targeting these brain areas. Second, there was a lack of recordings of responses to white noise – a major class of stimuli theorists use to probe and model information coding in the brain. Our project presented white noise stimuli. Together these data enabled us to test an old theory for the first time: that precise spike timing in response to white noise could be used like a barcode to classify neuron types.

Our project studies the transmission of visual information from the thalamus to cortex and to the hippocampus. We performed experiments and will present analysis demonstrating cortex like, sensory evoked responses in the hippocampus to visual motion, without any task, navigational or memory demands.

The OpenScope Databook: Reproducible System Neuroscience Notebooks to Facilitate Data Sharing and Collaborative Reuse with Open Science Datasets

Thanks for attending the webinar