Brain Observatory Toolbox

A MATLAB toolbox for accessing and using the neural recording public datasets from the Allen Brain Observatory¹. Available datasets:

• Visual Coding (Neuropixels and 2-Photon Calcium Imaging)
• Visual Behavior (Neuropixels and 2-Photon Calcium Imaging)

Contents

Getting Started • Going Further • Key Concepts • Installation • About • Acknowledgements

Getting Started

Quickstart examples for each Allen Brain Observatory dataset

Data releases from the Allen Brain Observatory include four datasets of neural activity recordings during presentations of visual stimuli to awake mice. Quickstart examples for each can be viewed, and readily run in MATLAB Online:

Dataset	Recordings	Experiment	Details	Quickstart Example
Visual Coding 2P 2	🔬 "ophys"a	Passivec	details	👀 (view) ▶️ (run)
Visual Coding Neuropixels 3	⚡ "ephys"b	Passivec	details	👀 (view) ▶️ (run)
Visual Behavior 2P 4	🔬 "ophys"a	Actived	details	👀 (view) ▶️ (run)
Visual Behavior Neuropixels 5	⚡ "ephys"b	Actived	details	(coming soon)

_{^a two-photon (2P) calcium imaging ^b large-scale neural probe recordings ^c presentation of various visual stimuli w/ untrained subjects ^d visual change detection task w/ trained subjects}

Technical white papers (see Details) provide information about the experimental protocols and computational pipelines for each dataset.

Programmatic workflow for accessing neural data in the Brain Observatory Toolbox

These three lines of code demonstrate the core workflow of the Brain Observatory Toolbox for accessing neural data, a workflow common to all datasets:

ophysSessionTable = bot.listSessions('VisualCoding', 'Ophys')
exampleSession = bot.getSessions( ophysSessionTable(1, :) )
dff = exampleSession.fluorescence_traces_dff

In the first line, listSessions lists available items (of type Session) from the Visual Coding 2P dataset. In the second line, getSessions creates an object to interact with the first listed item. The third line reterieves from the object fully preprocessed fluorescence traces representing the neural recording data. For further explanation, see Key Concepts and/or explore tutorials.

Going Further

Demonstrations illustrating neural data analysis

Demonstration examples illustrate neural data analysis concepts and practice while using the Brain Observatory Toolbox:

Dataset	Demonstration Example	About
Visual Coding 2P	👀 (view) ▶️ (run)	TODO
Visual Coding Neuropixels	👀 (view) (*)	TODO

_{(*) These data-intensive examples are currently recommended for use on local machines or user-configured cloud instances only, not for MATLAB Online}

Tutorials explaining access of dataset contents

Tutorial examples provide step-by-step guidance for using the Brain Observatory Toolbox to access and understand the metadata, processed data, and (where applicable) raw data available in the Allen Brain Observatory datasets:

Dataset	Demonstration Example	About
Visual Coding 2P	👀 (view) ▶️ (run)	TODO
Visual Coding Neuropixels	👀 (view) (*)	TODO

_{(*) These data-intensive examples are currently recommended for use on local machines or user-configured cloud instances only, not for MATLAB Online}

Key Concepts

Allen Brain Observatory dataset items

Allen Brain Observatory datasets each consist of various items of types depending on the specific dataset:

Brain Observatory Toolbox operations

The Brain Observatory Toolbox provides a set of two functions for each dataset item type:

• listXXX functions return a table of available items in the dataset
• getXXX functions return an object of selected item(s) by id(s) or table row(s)

This workflow is uniform across datasets and item types, and illustrated here:

The dataset item's contents are exposed through the object's properties and (where applicable) methods. For further guidance and explanation, see the Tutorials.

Installation

To install the Brain Observatory Toolbox persistently on a local machine or cloud instance, the Add-on Explorer is recommended:

1. Launch the Add-on Explorer
2. Search for "Brain Observatory Toolbox"
3. Press the "Add" button.

Required products

• MATLAB (R2023b or later)
• Image Processing Toolbox (if running the Visual Coding 2P demonstration VisualCoding_OphysDemo.mlx)

---

About the Brain Observatory Toolbox

🚧 The Brain Observatory Toolbox is at an early stage; the interface is not guaranteed stable across the v0.9.x releases.

💬 Questions? Suggestions? Roadblocks? Code contributors are regularly monitoring user-posted GitHub issues & the File Exchange discussion.

Acknowledgements

Initial engineering work, done by Ethan Meyers and Xinzhu Fang, was supported by the Foundation of Psychocultural Research and Sherman Fairchild Award at Hampshire College and hosted by the Center for Brains, Minds, and Machines at the Massachusetts Institute of Technology.

Footnotes

1. Copyright 2016 Allen Institute for Brain Science. Allen Brain Observatory. Available from: portal.brain-map.org/explore/circuits. ↩

2. Dataset: Allen Institute MindScope Program (2016). Allen Brain Observatory -- 2-photon Visual Coding [dataset]. Available from brain-map.org/explore/circuits. Primary publication: de Vries, S. E. J., Lecoq, J. A., Buice, M. A., et al. (2020). A large-scale standardized physiological survey reveals functional organization of the mouse visual cortex. Nature Neuroscience, 23, 138-151. https://doi.org/10.1038/s41593-019-0550-9 ↩

3. Dataset: Allen Institute MindScope Program (2019). Allen Brain Observatory -- Neuropixels Visual Coding [dataset]. Available from brain-map.org/explore/circuits. Primary publication: Siegle, J. H., Jia, X., Durand, S., et al. (2021). Survey of spiking in the mouse visual system reveals functional hierarchy. Nature, 592(7612), 86-92. https://doi.org/10.1038/s41586-020-03171-x ↩

4. Dataset: Allen Institute MindScope Program (2021). Allen Brain Observatory -- 2-photon Visual Behavior [dataset]. Available from brain-map.org/explore/circuits/visual-behavior-2p. ↩

5. Dataset: Allen Institute MindScope Program (2022). Allen Brain Observatory -- Neuropixels Visual Behavior [dataset]. ↩