DONNEES - Appel flash science ouverte : Pratiques de recherche et données ouvertes

A collective neuroscience laboratory: Beyond FAIR – NeuroWebLab

Submission summary

Biomedical research is transitioning towards increased openness and data sharing. The FAIR principles provide a framework to support this transition. By making the outcomes of research findable, accessible, interoperable and reproducible, it is expected to foster a more incremental science, where the results produced by one research team can seamlessly benefit the rest of the community. The challenge of open science and data sharing has strongly resonated in the neuroscience community: results from different fields had made clear that the challenge of the massive interindividual diversity will only be possible to tackle with large datasets. As a result, terabytes of neuroscientific data are openly available on the Web, including data for tens of thousands of individuals, spanning the scale of the molecule (whole-genome genotyping), the cell (histology), the complete organ (MRI, MEG, EEG), for the complete life span. However, the availability of this data has made the community realise that data sharing is just one part of the problem. To efficiently analyse and mine this data, scientists also need a methodology that would allow them to collaborate, working on the same data at the same time, instead of all redundantly and in parallel. Distributed collaboration paradigms have been well explored and adopted by the computer science community: tools such as Git and GitHub allow thousands of computer scientists to work on the same project, and consumer-grade tools such as Google Docs or the Wikipedia allow on a daily basis millions of users to work collaboratively. By contrast, the way in which most neuroscientists today work on shared data requires local download of the data, local processing, and eventually (although extremely rare), upload to a shared server. Even if data were findable, accessible, interoperable and reproducible, the process would still be inefficient, wasteful and cumbersome.
We propose to develop a Web platform – the NeuroWebLab – to go beyond FAIR, and allow neuroscientists to collaborate concurrently on the same data, extending the real-time paradigm of tools such as Google Docs to different data modalities. During the last years we have gathered extensive experience building Web tools for real-time collaboration: Brainspell, BrainBox and MicroDraw. We propose to build on this experience to develop a general Web platform for distributed scientific collaboration on open data, which will include the data modalities we are developing ourselves, and which could also be easily extended to new data modalities. During the 2 years of the project we propose to develop an open source Web platform that will combine: (1) collaborative editing of metadata, (2) indexation of the neuroimaging literature, (3) a curation, visualisation and editing of brain imaging data, (4) curation, visualisation and editing of histological data, (5) a framework for the creation and administration of collaborative projects of distributed research teams, (6) an application programming interface (API) facilitating the access to the data by data scientists.
Our development will adhere to open science best-practices, including standard project structures and data formats, and will be extensively documented and released open source.

Project coordination

Roberto Toro (INSTITUT PASTEUR)

The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.

Partner

JOGL Just One Giant Laboratory
IP INSTITUT PASTEUR

Help of the ANR 100,000 euros
Beginning and duration of the scientific project: October 2019 - 24 Months

Useful links

Explorez notre base de projets financés

 

 

ANR makes available its datasets on funded projects, click here to find more.

Sign up for the latest news:
Subscribe to our newsletter