testingGIN/minutes/2019-08-08-matthewmeeting.md

Meeting outputs

main decisions

• force19, ok
• continue discussion with other SFB, info dissemination project is our to lead
• Get a meeting with Thomas. We do not care who develop it but we need something going to start testing.
• Continue to work on mice metadata (some files will be sent by matthew soonish)

strategy

Main objective is information dissemination, the development of the tool should keep this in mind.

• Get MVP with commenting
• Analyse comment to find gaps
• modify tool to push people to fill the gaps (empty box, badges,...)

software

• need simplified workflow, only one tab ?

visualisation

• need search filter function
• interaction with author should be straightforward
• maybe need to link to something they already know, as a reminder of what we are talking about.

My role

• ASAP stop developing, but foster tool use
• social component

#Vorbereitung

1. Collaborations

• Thomas wants to clarify goals and work together (why we do OUR software)

• possible interaction with other SFB: contact with sfb1158 in heidelberg (see below: metadata standards)

• proposition *

  • Two software is good: interoperability of what is created is easier to demonstrate, our will not be scalable but will be easier to set up.
  • Need to think about a decentralised system, i.e. look into https://linkedresearch.org community ?
  • Need to work on the smartfigure xml standard: this needs more engagement and an open developemnt, sourcedata should lead that in collaboration with elife/texture people.
    

2. Shiny Software

• dropbox version
• server version with ITB server

Visualisation and data upload in a 2 tab browsable software

3. Visualisation

• highlighted versus normal entries work
• google-like visualisation has limitations
• NEED visualisation of protocols ?
• long term: browsability of results

** Outside of my area of expertise **

4. Standards

• Smartfigure: zip +xml
• metadata: minimal standards:
  • animal (species, genotype, date_of_birth, light modus, diet) - similar to/extended from pyrat ?
  • experiment (date, time, temperature, light modus)
• https://www.nwb.org/ standard for data (python, matlab export, called hdf5)
  
  

5. RDM work

• talk with Robert, work with keisuke:
  • data of paper to be organised, pushed to GIN
  • structure worked, file naming comes next.
  • time to perform the work will be reported : paper to follow when 3-4 case studies are there.
  • next steps will be to share the analysis code.
• next will talk with christian for imaging data, using same approach.
  

6. survey and server

• 14 labs responded
• estimation gots from 5 to 200 TB (whole lab data)
• need to talk to itb people about it.