Data-model and file structure for OMERO.editor files

[will]

Several requirements:

Need to facilitate use of editor files by other software. eg: Excell can read MAGE-TAB files, or html is read by browsers etc.

Need to support multiple parameters per field (new data structure)

Would be nice to have parameters in context (within text).

Current files are very hierarchical. Aim to support this (and support hierarchy in future).

Must support tabular data.

Should support imported XML. Save as unaltered XML?

Need to save Editor files locally, e-mail them etc.

How should Editor files be stored on the OMERO server?

MAGE-TAB

MAGE-TAB is one possibility: see http://www.mged.org/mage-tab/ for paper etc. Some example files here: http://tab2mage.sourceforge.net/docs/magetab_subs.html MAGE-TAB uses several tab files to describe Microarray data. Closest to what we are interested in is the IDF (investigation design format), which describes the protocols used, with names/id, description and any parameters. These protocols are referred to by the Sample and Data Relationship Format (SDRF) spreadsheet, which describes a DAG of the workflow.

The tab format is used by MAGE for several reasons:

• Easy to represent a DAG, especially since layers of the DAG are usually of the same type of object.
• DAG nodes have small number (1-3) inputs / outputs
• Biologists are familiar with using spreadsheets to store data.

"Semantic" wiki

Could use a wiki to edit text in a familiar manner, while marking up the experimental variables in some way? Manually, e.g. Resuspend cells in {{Parameter 10 ml}} media. Or with Toolbar buttons?

XHTML

This might address most requirements? Could use <div> elements, <p> elements <table> elements etc.

Presented the above ideas at Client meeting, 8th Oct'08. http://users.openmicroscopy.org.uk/~will/presentations/

After discussion, conclusion was to use XHTML as format for OMERO.editor, but also support import and export to text and MAGE-TAB based formats.