NDI: A platform-independent data interface and database for neuroscience physiology and imaging experiments

eNeuro. 2022 Jan 21:ENEURO.0073-21.2022. doi: 10.1523/ENEURO.0073-21.2022. Online ahead of print.

ABSTRACT

Collaboration in neuroscience is impeded by the difficulty of sharing primary data, results, and software across labs. Here we introduce Neuroscience Data Interface (NDI), a platform-independent standard that allows an analyst to use and create software that functions independently from the format of the raw data or the manner in which the data is organized into files. The interface is rooted in a simple vocabulary that describes common apparatus and storage devices used in neuroscience experiments. Results of analyses - and analyses of analyses - are stored as documents in a scalable, queryable database that stores the relationships and history among the experiment elements and documents. The interface allows the development of an application ecosystem where applications can focus on calculation rather than data format or organization. This tool can be used by individual labs to exchange and analyze data, and it can serve to curate neuroscience data for searchable archives.Significance StatementNeuroscience experiments generate heterogeneous data, and each lab typically stores its data and analyses in their own idiosyncratic formats and organizations. We introduce an interface standard - the Neuroscience Data Interface - that allows the user to specify these formats and organizations so that data and analyses can easily be shared among labs or posted to journals and archives.

PMID:35074827 | DOI:10.1523/ENEURO.0073-21.2022