Allen Institute and Neurodata without Borders (Nicholas Cain, AIBS)

Using the Allen Institute’s Brain Observatory to Obtain Neural Data Without Borders

Nicholas Cain introduced a great local resource in obtaining freely available neural data— The Allen Institute for Brain Science. More specifically, he illuminated aspects of its Brain Observatory that contains a plethora of 2-photon calcium imaging maps of mouse visual cortex, all of which have been painstakingly pre-processed and made accessible by researchers at AIBS.

Before diving into the nitty-gritty of how to obtain files from the Brain Observatory, Nick gave a great overall introduction to the Allen Institute. It is a non-profit centered on basic science research located here in Seattle. One of the major focuses of the institute is to bridge gaps between different scales of neuroscience, such as explaining cognition at the circuit-interaction level. Nick’s role in furthering this goal is to work to make data as easy as possible to obtain for extramural researchers to analyze.

As mentioned earlier, The Brain Observatory consists of a massive database of calcium imaging maps obtained from awake mice. The mice are not engaged in goal-directed behavior but instead passively view a myriad of static and dynamic images such as natural scenes and drifting gratings. Altogether the Observatory’s website contains about 38,000 isolated cells and responses for contributors to further analyze. Characterized cells’ responses can be easily viewed by navigating to the Cells page of the observatory website. Here information such as orientation tuning and responses to particular stimuli is displayed as various graphs made to clearly elucidate the cells preferred stimuli across all those shown.

Nick also highlighted the key points in obtaining the imaging data for further analysis. The data comes in a specific Neurodata Without Borders format, which is a type of hierarchical file format (HDF5; that can be accessed through the HDFview software. This requires a pip install of the Allen Software Development Kit using The Jupyter Notebook. The brain observatory’s cache can be accessed via the Allen SDK, and obtaining experimental data simply requires creating a cache object & pinging the network to produce the data. Getting specific experiments is then a matter of locating the ophys ID of interest and downloading that NWB within the SDK. Several comprehensive tutorials are available for review within the Allen SDK’s website.