Neurodata without Borders Export

[bartkrekelberg]

I am revisiting the export() utility (#144) and planning to make something to export to Neurodata Without Borders format. The BRAIN initiative is pushing this hdf5 based format in its data sharing repositories, it is cross-platform (Pyhthon/Matlab), and extensible.

This would not replace saving the CIC object (#81), but the goal is that the NWB file stores everything needed for data analysis (and provenance tracking) so that other researchers can use their NWB tools to re-analyze Neurostim data.

I plan to start with generic items (e.g. plugin properties, analogous to the current export() branch) and then move on to exports of the external data (e.g. eyelink, ripple).

If anyone has already tried something like this, please let me know.

[adammorrissirrommada]

This sounds excellent. There were some discussions about this at our end too, but no concrete steps. @cnuahs, @nicholasprice, @ezavitz are interested I think. They're not using Ripple, but can hopefully replicate your solution for their systems (e.g. Blackrock).

[nicholasprice]

I'd support efforts for exporting OpenEphysGui data. Not Blackrock though.

[adammorrissirrommada]

@nicholasprice, does anyone at your end have time to look into this, if it is still relevant to you? there is a stale pull request from @bartkrekelberg. I think one of the challenges with that PR is that it seems the entire cic.m comes up as changed in the diff, at least on GutHub.

[nicholasprice]

I had a look through the commits, but don't have the neurostim-nous to review this. @ezavitz might have the knowledge, but she doesn't have the time! n

On Thu, Jan 27, 2022 at 9:53 AM Adam Morris @.***> wrote:

@nicholasprice https://github.com/nicholasprice, does anyone at your end have time to look into this, if it is still relevant to you? there is a stale pull request https://github.com/klabhub/neurostim/pull/144 from @bartkrekelberg https://github.com/bartkrekelberg. I think one of the challenges with that PR is that it seems the entire cic.m comes up as changed in the diff, at least on GutHub.

— Reply to this email directly, view it on GitHub https://github.com/klabhub/neurostim/issues/170#issuecomment-1022676456, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHFMZC2JMB6BQQUBCNDSVR3UYB3OLANCNFSM46Y77UHA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you were mentioned.Message ID: @.***>

-- A/Prof Nicholas Price Phone: +61 3 9905 5131 / 0424 56 14 17

[ezavitz]

Sorry @adammorrissirrommada - I have neither the knowledge nor the time unfortunately! I support the idea of the NWB format, and I think that is the way I want to go with archival data as well, but will need to get involved with testing when I can integrate it into something I'm working on and have re-learned what I was doing before. Do you want me to follow up with you when I get to that point?

[adammorrissirrommada]

I am agnostic - I am out of the game mostly, so won't be using it myself - so we can leave it as a stale PR until someone feels the need to nudge it along. Thanks for the update.

[bartkrekelberg]

Just a note that this is still on my radar. As archival/shareable data format NWB seems to be the way to go, but not everything in CIC may be stored. (at least that is not my priority at the moment).

I am also integrating this with a datajoint pipeline for Ns. Currently, this is on the NWB branch of the neurostim repo in an early alpha version, but maybe this should be its own repository.