AnnoJ won't 'syndicate'

[danmaclean]

AnnoJ won't syndicate because the config/config.yml is wrong for this app. public/javascripts/config.js is read by AnnoJ on start to find out where its resources are, config/config.yml is an easy way (well for my users who get annoyed by JSON) to specify the information for config.js. Running rake create:config generates a config.js from the yml. Here is an example config.yml for two tracks, the gene features and some reads.

tracks:

• id: 1 name: My Gene Models Track type: ModelsTrack path: Gene Models data: /features/annoj/1 showControls: true height: 80
• id: 2 name: Reads from an RDBMS provided experiment loaded via gff type: ReadsTrack path: Reads data: /features/annoj/2 showControls: true minHeight: 20 maxHeight: 500 height: 70

  A list of tracks that will be active by default (use the ID of the track)

  active:

  • 1
  • 2

    genome - the service that knows about the genome

    genome: genomes/annoj/1

    bookmarks - the service that keeps the bookmarks

    bookmarks:

    css - a list of services or locations for css (optional)

    stylesheets:

    • id: simple name: simple CSS href: stylesheets/simple.css active: true

      where to focus the view on start up

      location: assembly: Chr1 position: 4000 bases: 20 pixels: 1

Fixing this makes AnnoJ play nicely once the genome and features are added through the interface

[mrship]

I can't get the YML file to parse if I copy the above straight into config.yml and various attempts at putting into what appear to be a valid YML file don't work either. Can you attach the version you've got running instead?

[danmaclean]

The file is at https://gist.github.com/danmaclean/5355427 , hope this works

[mrship]

That parses to allow for the config.js to be created, but under when I browse, under System Messages I get

ActiveRecord::RecordNotFound in FeaturesController#annoj
Couldn't find Experiment with id=2

I'm not sure whether I need to get this working or not though, so if you're happy with it on your end we can move on.

[danmaclean]

Now you need to add data to the database, see the manual rbrowser.pdf pages 8, 9, 10. Add a new genome with the data in public/sequences/sample_reference_TAIR9_chr1.fna, add a feature set with the data in public/sample_gffs/sample_features.gff, then add a read feature set public/sample_gffs/sample_reads.gff. This should create the reference and gene tracks and some reads you specified in that yaml. Otherwise delete track 2 from the yaml and regenerate config.js.

AnnoJ should then have something to read from the DB.

But yeah, I can test this and it seems like the sort of error we should get if the error is the DB isn't populated. Basically as a minimum you need the genome and some gene features in there for anything to display.