Cytoscape 2.3 tasks brainstorm
Right now, these are just Gary's notes for discussion purposes
Large
- New drawing library (Nerius)
- Refactor Cytoscape class
MetaNodes, Hyperedge API + visualization options
- API to deal with synonyms, identifiers and URL links to external sources - a node ID and synonym unification API - this can be populated from files or via a cPath web service
Medium
Reimplement ExpressionData object as DataMatrix on top of CyAttributes
- Refactor initialization to be more general (e.g. load any file from local disk or URL) - will help with webstarts. Review all command line/cytoscape.props arguments and clean up where required.
- Clean up attributes file - revisit Rowan's new attribute file as the default attribute file type to "import" from - keep old parsers as import from old attribute file types menu option.
- Remove all default node and edge attribute names (canonical name, aliases, etc.). Only default attribute should be ID - create a simple architecture for pluging in attribute semantics on an as needed basis. A basic semantics package could have a label attribute.
- Replace select nodes by name (ctrl-f) with a mozilla like find service - type in any regex and all matching node labels are selected. Filters can do this, but we need something blindingly simple for common uses.
- Move all user documentation to the wiki and use this to generate a PDF and Java help
Small
- Clean up Cytoscape classes, directory, library files, package names (all API classes should start with Cy, all application specific classes should start with Cytoscape)
- Addition of generic link out plugin to the core - separation of yeast-context into 3 plugins - one for graphical context menus, one for general links e.g. to Google, one for organism specific links (the generic link out plugin)
- Ability to rename networks in the network manager