Why?

Microscopy data is big and complex

• Long experiments
• Screens (maybe not here)
• Data intensive microscopes (e.g. Light sheet)
• Complicated data (super-res, speckle)

Keeping data accessible

(for you and others)

• A file browser is not a data management tool
• Enforcing standards within your group can be hard
• What happens when students / post-doc are gone?
• Collaboration with data-scientist can be a challenge

Data management plan

• Mandated by institutions or the ANR (since last year)
• As open as possible, as closed as necessary
• Here is a template

FAIR

Findable

Accessible

Interoperable

Re-usable data

What?

Some history

• 1990s first commercial CCD
• 2000-2010 the Metamorph era (and nd / stk files)
• 2005 sq. formats explosion (vendor lock-in strategy)
• since 2010 :
  • change of paradigm regarding open-source,
  • federation of global microscopy community

Openmicroscopy provides standards

• Managed by the U. of Dundee group / Glencoe software (Jason Swedlow, Josh Moore)
• Defined OME-TIFF (data + metadata in a single file)
• created BioFormats
• Omero is also used in the industry (CROs, Perkin Elmer)

Dataflow

Data Federation

At the University, National and International levels

Features overview

How and when?

Costs

• One server (less than 3k€)
• Mainly HR

HR implications

• Research engineer @ 50% FTE (centuri):

  • manage the federated databases
  • interface with data analysts
  • custom dev

When?

Setup step (first trimester 2021)

• Renew list of personnel, access to disks
• Automated import strategy (old data)
• Automated import strategy (new data)
• Automated annotation ?

Usage and adoption (throughout 2021)

• Training of post-docs & PhDs
• Freezing of the production workflow