NWB Query Engines: Tools to Search Data Stored in Neurodata Without Borders Format
Petr Ježek1, Jeffery L Teeters2, Friedrich T Sommer2
1Faculty of Applied Sciences, New Technologies for the Information Society, University of West Bohemia, Plzeň, Czechia.
Frontiers in Neuroinformatics
|October 12, 2020
Summary
New tools enable searching neurophysiology data stored in Neurodata Without Borders (NWB) files. These solutions facilitate efficient discovery of specific NWB datasets for research purposes.
Area of Science:
- Neuroscience
- Data Science
- Computational Biology
Background:
- The Neurodata Without Borders (NWB) format is a standard for storing neurophysiology data and metadata.
- Current limitations exist in searching and retrieving specific data from large collections of NWB files.
Purpose of the Study:
- To introduce and describe three novel tools designed for searching collections of NWB files.
- To provide researchers with efficient methods for data discovery within NWB archives.
Main Methods:
- Development of the NWB Query Engine (Java) for comprehensive NWB file content searching.
- Implementation of 'search_nwb' (Python) for searching both NWB 1 and NWB 2 file contents.
- Creation of 'nwbindexer' utilizing a two-step SQLite database indexing process for faster, albeit less comprehensive, searches.
Main Results:
- Three distinct tools (NWB Query Engine, search_nwb, nwbindexer) offer varied functionalities for NWB data searching.
- The tools support a custom query language and are integrated into a web interface for user accessibility.
- NWB Query Engine offers speed for certain queries; nwbindexer provides faster searches after indexing.
Conclusions:
- The developed tools address the critical need for efficient search capabilities within NWB data repositories.
- These tools enhance data accessibility and facilitate the reuse of neurophysiology data for scientific advancement.
