LogoLogo

NWB — Neurodata Without Borders

Overview

NWB (Neurodata Without Borders) is the community standard for neurophysiology data, covering intracellular and extracellular electrophysiology, calcium imaging, optogenetics, and behavioural data. Initiated in 2015 with funding from the NIH BRAIN Initiative and co-founded with the Allen Institute for Brain Science, NWB addresses the absence of a shared data format that made cross-lab reuse and meta-analysis of electrophysiology data difficult. It uses HDF5 (and optionally Zarr) as its underlying container format, with a rigorously defined schema organising neural data and all associated metadata into a single self-describing file. It is governed by the NWB Working Group and is the mandatory format for the DANDI Archive.

Data Types

NWB handles the full range of neurophysiology and systems neuroscience data, including extracellular electrophysiology (spike-sorted unit data, raw LFP, electrode geometry), intracellular electrophysiology (patch-clamp in current and voltage clamp), calcium imaging (two-photon and wide-field fluorescence, ROI segmentations, ΔF/F traces), optogenetics (stimulus parameters and light delivery metadata), and behavioural data (positional tracking, task events, stimuli, reward delivery).

NWB Data Extensions

The NDX (NWB Data Extension) mechanism allows community-developed extensions to the core NWB schema for lab-specific or modality-specific data types. Extensions are versioned Python packages published on PyPI and catalogued at https://nwb-extensions.github.io.

Relationship to BIDS

NWB and BIDS serve complementary roles. BIDS organises datasets at the file and directory level across modalities, while NWB is a rich single-file format for neurophysiology recording sessions. BIDS has a dedicated extension proposal (BEP032) for organising NWB files within BIDS structure, and the DANDI Archive supports both formats.

Connections

Resources

Graph View

Tags

Backlinks