Open Data In Neuroscience (ODIN) Initiative

The Open Data in Neuroscience Initiative at the McGovern Institute for Brain Research at MIT is committed to democratizing neuroscience research by cultivating an open data ecosystem both within and outside MIT. We aim to provide neuroscientists with robust platforms and tools to share, analyze, and interpret data, while synchronizing with other open data initiatives and federally funded projects. Through a collaborative and transparent approach, we strive to accelerate the discovery and understanding of neural systems and their impact on human health and behavior.

The initiative will work synergistically with established initiatives such as DANDI, BRAIN Initiative Cell Atlas Network, BRAIN CONNECTS, and others. Our goal is to foster a community where data sharing becomes the norm, not the exception, thereby encouraging scientists to work together to unravel the complexities of the brain in ways that isolated efforts cannot achieve. In turn, these efforts will integrate ethical considerations, reflect diverse perspectives, and sustainable stewardship of neuroscience resources.

The initiative will co-develop toolkits that enhance rigor and reproducibility through the integration of open neuroscience data archives, open ML/AI models, and use of standards and best practices. These toolkits can be embedded into classroom settings alongside problem sets and other resources to address current gaps in curricula, and disseminated to the neuroscience community for adapting and adopting into different settings. Annual workshops will complement these approaches by bringing thought leaders to foster interdisciplinary collaborations, share insights, and integrate tools and techniques that can benefit neuroscience research.

We aspire to push the boundaries of what is scientifically possible in understanding the neural basis of human cognition, behavior, and disease. We invite all neuroscientists, data scientists, and stakeholders committed to this approach to join us in this transformative journey.

DANDI

DANDI (Distributed Archives for Neurophysiology Data Integration) is a web platform for scientists to share, collaborate, and process data from cellular neurophysiology experiments. DANDI works with the BRAIN Initiative and other awardees to curate data using community data standards such as NWB and BIDS, and to make data and software for cellular neurophysiology FAIR (Findable, Accessible, Interoperable, and Reusable).

The DANDI Archive currently hosts over 350 datasets (580TB), including raw and processed data, and associated software containers. The data archive provides a broad range of cellular neurophysiology data, including electrode and optical recordings, and associated imaging data using a set of community standards: NWB-Neurophysiology, BIDS - Brain Imaging Data Structure, and NIDM - Neuro Imaging Data Model.

DANDI provides a number of features to facilitate data sharing and collaboration, including:

• A user-friendly web interface for searching and browsing datasets
• A JupyterHub deployment for interactive data analysis
• A command-line interface for programmatic access to datasets
• A community forum for discussion and support