nwb4edu: Teaching and Learning with NWB Datasets

An interactive online textbook for teaching and learning neuroscience using Neurodata Without Borders (NWB) datasets.

Website: https://nwb4edu.github.io

JOSE Paper:

About This Project

The nwb4edu online textbook provides interactive and self-guided resources for educators, students, and self-guided learners who would like to use open neuroscience datasets for research and/or teaching. The textbook introduces users to Neurodata Without Borders (NWB), a data standard for neurophysiology data, and teaches how to access, analyze, and visualize NWB datasets using Python.

Key Features

• Interactive Jupyter notebooks with executable code
• Comprehensive tutorials covering data access, analysis, and visualization
• Problem sets for hands-on practice
• Real datasets from DANDI Archive and the Allen Institute for Brain Science
• Greatest Hits lessons that recreate figures from landmark neuroscience papers
• Launch on JupyterHub - no local installation required

Content Overview

The textbook is organized into several sections:

Introduction

• Using this book
• Environment setup
• Resources for educators

Data Science in Python (Refresher)

• NumPy
• Pandas
• Matplotlib
• SciPy

PyNWB Lessons

Lesson 1: Introduction to NWB format

• Obtaining datasets with DANDI
• Working with NWB format in Python
• Interactive NWB data exploration

Lesson 2: Large-scale electrophysiology

• Spike sorting
• Visualizing extracellular recordings

AllenSDK Lessons

Lesson 3: Single-cell electrophysiology

• Obtaining data for single cells
• Analyzing computed features

Lesson 4: Two-photon calcium imaging

• Retrieving Brain Observatory data
• Analyzing two-photon data
• Correlations in neural activity

Greatest Hits in Neuroscience

Recreate figures from foundational papers:

• Place fields (O'Keefe & Dostrovsky, 1971; Grosmark & Buzsáki, 2016)
• Rotational dynamics in motor cortex (Churchland et al., 2012)

Using This Textbook

For Students

The easiest way to use this textbook is through the live website at https://nwb4edu.github.io. You can launch any notebook directly on JupyterHub using the rocket icon in the top right corner of each page. No installation on your computer is required when using JupyterHub!

For Local Development

If you want to run the notebooks locally:

1. Clone this repository:

git clone https://github.com/nwb4edu/nwb4edu.github.io.git
cd nwb4edu.github.io

2. Install dependencies:

pip install -r requirements.txt

3. Launch Jupyter:

jupyter notebook

For Educators

This textbook is designed to be modular - you can use individual lessons or entire sections in your courses. All materials are licensed under CC-BY-SA-4.0, so you're free to adapt and modify them for your teaching needs.

See our For Educators guide for more information on integrating these materials into your curriculum.

Building the Book

This textbook is built using Jupyter Book.

Requirements

Important: Building the book requires Python 3.11 due to AllenSDK compatibility requirements (AllenSDK requires numpy<1.24, which is incompatible with Python 3.12+).

1. Create a Python 3.11 environment:

conda create --name nwb4edu python=3.11 -y
conda activate nwb4edu

2. Install dependencies:

pip install -r requirements.txt
pip install -r requirements-docs.txt

Build Process

To build the book locally:

jupyter-book build .

Or use the provided Makefile:

make build

To clean the build directory:

make clean

The built HTML files will be in _build/html/.

Deploying to GitHub Pages

The book is automatically deployed to GitHub Pages when changes are pushed to the master branch. To manually deploy:

make deploy

This will build the book and push the HTML to the gh-pages branch.

Repository Structure

nwb4edu.github.io/
├── _config.yml              # Jupyter Book configuration
├── _toc.yml                 # Table of contents
├── landing_page.ipynb       # Home page
├── Introduction/            # Getting started materials
├── Data_Science_In_Python/  # Python refresher lessons
├── Lesson_1/               # NWB format basics
├── Lesson_2/               # Electrophysiology analysis
├── Lesson_3/               # Single-cell data
├── Lesson_4/               # Two-photon imaging
├── Lesson_5/               # Greatest Hits recreations
├── requirements.txt         # Python package dependencies
└── requirements-docs.txt    # Documentation build dependencies

Contributing

We welcome contributions from the community! Regardless of whether you want to report a bug, typo, or issue, suggest improvements to existing lessons, contribute new lesson content, or improve documentation, please see our Contributing Guidelines for more information.

Getting Help

• General questions: Open a discussion
• Bug reports: Open an issue
• Direct contact: Email Ashley Juavinett at ajuavine(at)ucsd(dot)edu

Authors

• Ashley L. Juavinett - UC San Diego, Neurobiology (current project maintainer)
• Victor Magdaleno-Garcia - UC San Diego Alum, Cognitive Science

With thanks to Tommy Lee, Ariel Rokem, and Tom Donoghue for copyediting.

Acknowledgements

This work is supported by a Kavli Foundation Seed Grant to increase access to Neurodata Without Borders datasets. We thank the Kavli Foundation, specifically program officer Stephanie Albin, for their support.

We also thank Ben Dichter, Oliver Rübel, and Ryan Ly for valuable feedback and technical support.

License

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

You are free to use and adapt these resources, but if you create derivative works, they should be distributed under the same license.

Citation

If you use these materials in your teaching or research, please cite:

Juavinett, A.L., & Magdaleno-Garcia, V. (2024). nwb4edu: an Online Textbook for Teaching
and Learning with NWB Datasets. *Journal of Open Source Education: TENTATIVE*.
https://nwb4edu.github.io