Lorentz Center - Share and Flourish: New Standards for Data Sharing in the Neurosciences from 18 Aug 2014 through 22 Aug 2014
  Current Workshop  |   Overview   Back  |   Home   |   Search   |     

    Share and Flourish: New Standards for Data Sharing in the Neurosciences
    from 18 Aug 2014 through 22 Aug 2014


Share and Flourish: New Standards for Data Sharing in the Neurosciences

Lorentz Center, 18-22 Aug 2014
Rembrandt Bakker, Moniek Lijster, Nick Ramsey, Paul Tiesinga, Fons Verbeek, Hugo Vrenken, Tonya White

Neuroscience has great potential to answer one of the big questions in life: How does the brain work? More than hundred thousand researchers are working in this field, but the experimental data that is acquired and published about is often not available. Given that progress in the neurosciences is likely to come from the integration of data across scales and modalities, the lack of data sharing is not acceptible. In this workshop we have considered several options to break through the barriers to data sharing:

1.     Create a social environment in which data hiding is considered bad scientific practise

2.     Making data sharing count in similar ways as publications count

3.     Making data sharing mandatory for research done with public money

4.     Show examples of how sharing data makes a researcher successful

5.     Take away the cost of datasharing for the individual researcher, by organizing it at the institutional level

6.     Create infrastructure that enables researchers to find data based on textual, regional or coordinate-based queries

The workshop had 33 Dutch, 11 non-Dutch European, and 12 overseas participants, and aimed to produce a deliverable in the form of a White Paper, for which a starter document had been prepared by the organizers. Discussions at the workshop were intense but always constructive, and a set of thirty recommendations were presented by Paul Tiesinga at the INCF Congress in the week after. Some highlights per topic:

Ethics – Much restraint to share human brain data is caused by trying to be “better safe than sorry” when it comes to privacy laws. Authoritative guidelines can help.
Credits – The Allen Institute has demonstrated that the release of well documented data is like planting a tree: you will bear fruit for many years to come.
Integrity – Data with minimal meta data (‘Dublin core’) has little value except for methodological research.
Standardization – If standard brain atlases cannot be agreed on, then at least we need practical tools to convert between them.
Sustainability – Repositories at the institute level are preferred, given the expertise required to make data ready for integration.

A writing team with ten active members is going through the huge body of reference material that we gathered on a shared drive, to convert the draft white paper into a publishable manuscript.

The format of the workshop was a pre-scheduled mixture of talks, panel discussions, break-out sessions and lunch breaks, which could have been half an hour longer than the two hours alotted. Key-note speakers typically stayed a few days rather than the full week, but younger scientists and developers were very devoted. An eye opener was the quote from Aad van der Lugt after the presentation of recommendations. After presenting the Rotterdam scan study that is accessible only through paid subscription, he answered the question whether all publicly funded neuroscience data should be freely available with a clear ‘Yes, it should!’.