Towards open and reproducible neuroscience in the age of big data
Download as PPTX, PDF1 like645 views
This document discusses open data sharing in neuroscience. It begins by introducing NeuroVault, a platform for sharing neuroimaging data. It then discusses various incentives for data sharing, such as journal policies requiring data availability and opportunities for data papers. Standards like BIDS are presented as a way to make data easier to use. OpenNeuro is highlighted as a new platform for sharing preprocessed neuroimaging data and running analyses. The benefits of open data are discussed, such as enabling faster, cheaper, higher quality, and more inclusive/competitive research. The conclusion advocates getting more data into researchers' hands and building tools to achieve this goal.
Towards open and reproducible neuroscience in the age of big data
- 1. Towards open and reproducible neuroscience in the age of big data Chris Gorgolewski @ChrisFiloG http://bit.ly/2ifdaeX
- 2. Who discovered the structure of the DNA?
- 6. Revisiting the Unreasonable Effectiveness of Data https://arxiv.org/abs/1707.02968
- 7. Maximizing the impact Getting more data out there
- 9. Neuroimaging data sharing hierarchy Poldrack and Gorgolewski, 2014
- 11. NeuroVault stats • 35,000 maps • 1440 collections • 61% public • 21% linked • Representing 86 different journals • 296 mentions on Google Scholar
- 12. Ultimate consent form • Inform participants about your intention to share data • Explain the benefits • Discuss the risks open-brain-consent.readthedocs.org
- 13. Incentive I - policies
- 14. Journal policies – case of NPG Checklist with compulsory data availability declaration
- 15. “Badges are stupid, but they work.” – Brian Nosek
- 16. Journal policies – case of PloS “All data and related metadata underlying the findings reported in a submitted manuscript should be deposited in an appropriate public repository, unless already provided as part of the submitted article”
- 17. Incentive II - credit
- 18. Gorgolewski, Milham, and Margulies,
- 19. • Neuroinformatics (Springer) • GigaScience (BGI, BioMed Central) • Scientific Data (Nature Publising Group) • F1000Research (Faculty of 1000) • Data in Brief (Elsevier) • Journal of Open Psychology Data (Ubiquity press) • Frontiers Where to publish data papers?
- 20. Incentive III - analysis
- 21. NeuroVault benefits - decoding
- 22. NeuroVault benefits – similarity search
- 23. NeuroVault benefits – similarity search
- 24. NeuroVault benefits – Gene decoding Gorgolewski KJ, Fox AS, Chang L et al. Tight fitting genes: finding relations between statistical maps and gene expression patterns. F1000Posters 2014,5:1607 (poster)
- 25. NeuroVault benefits - Gene decoding
- 26. a free online platform for sharing and analysis of neuroimaging data
- 27. OpenNeuro.org - Poster #1677 28
- 28. Demo OpenNeuro.org - Poster #1677 29
- 29. OpenNeuro - available Pipelines • FMRIPREP • MRIQC • C-PAC • FreeSurfer • ndmg • SPM • BARACUS • MAGeTBrain Coming soon: • QAP • OPPNI • automatic analysis • TRACULA • BROCCOLI 30
- 30. MRIQC – Quality control for structural and functional images MRIQC.ORG
- 33. OpenNeuro is free to use by anyone under the agreement that the data will be made publicly available after 18 months.
- 34. Reproducibility Data snapshots + software containers == reproducibility
- 35. Maximizing the impact II Making data easier to use
- 36. Brain Imaging Data Structure bids.neuroimaging.io
- 37. OpenNeuro - Preprocessed data
- 40. Science in the age of Open Data
- 41. Faster Going from hypothesis to answer in a couple of weeks
- 42. Cheaper ~$3 mln cost of reacquiring data for each of the reuses of OpenfMRI datasets (2017)
- 43. Higher quality Wicherts JM, Bakker M, Molenaar D (2011) Willingness to Share Research Data Is Related to the Strength of the Evidence and the Quality of Reporting of Statistical Results. PLoS ONE 6(11): e26828. doi: 10.1371/journal.pone.0026828
- 44. More inclusive and competitive Same datasets are available to all researchers.
- 45. Conclusion 1. we need to get more data in the hands of more researchers 2. building tools and platforms is crucial to achieve this goal
- 46. Poldracklab Know of any neuroinformatics faculty jobs? Get in touch! (asking for a friend ;)