Documentation of ImageTrack JSON format

[jbrowsetix]

So people can write their own renderers.

People love doing this. Love it. Hacking together your own glyphs for GBrowse is so popular that there was a massive epidemic of it in the 80’s and Nancy Reagan had to go on TV saying that the current glyph set was more than adequate.

original LH ticket

This ticket has 2 attachment(s).

[jbrowsetix]

Documentation of ImageTrack JSON format

Just downloaded and started trying to set up JBrowse, and immediately wanted to try writing my own -to-json.pl script since my source data is in other formats (GBK, BSML, etc.) than those listed on the tutorial docs. I’ll work on it by looking at the example data files and code, but having everything written up would help. (I know, I know, when I figure it all out *I can write it.)

by Joshua Orvis

[jbrowsetix]

Documentation of ImageTrack JSON format

Hi guys

Almost a year on since this ticket.

Yes, indeed documentation for JBrowse is scarce which is surprising for something published in Genome Res a year ago.
I don’t think making your own glyphs is a big problem currently, as I couldn’t find basic documentation (perhaps I am stupid or not looking at the right place?). As an example, the only documentation I could find for -db_args in the primary configuration file is this:

db_args | arguments to the constructor for the specified db_adaptor | dictionary of key-value pairs

Likewise, the perldocs are equally non existent. Apologies if this is existing somewhere but I couldn’t find in the docs/ or the biowiki (there is some but i wouldn’t find it useful for developing - it was very useful, however, as an explanation of how it does what it does...).

But no, I don’t think that community users like Joshua or myself should write this initial documentation because we can easily get it wrong.

We would really benefit from the having access to your JSON specs and more documentation for the configuration files. Having more documentation would help us who might be interested in developing jbrowse instead of the point-and-click replacement for GBrowse (which is very well documented). And I have to say I’m excited from the little I’ve seen about Jbrowse (relying on an external library like dojo was a great idea) and currently the documentation is the main bottleneck.

Once the main docs are available then we would be happy to help extending and improving it!

thanks
a

by Alexie Papanicolaou

[jbrowsetix]

Documentation of ImageTrack JSON format

p.s. Yes, perldoc provides the arguments needed but it was just first example as I went down the config-desc.html. I’m comparing with GBrowse’s documentation (which one can start copy paste in order to make the JBrowse one) @ http://gmod.org/wiki/GBrowse_Configuration_HOWTO

by Alexie Papanicolaou

[jbrowsetix]

Documentation of ImageTrack JSON format

I guess it would be better if I wrote what would be cool to have. Attached is how far I got today. The image is a JBrowse of data generated on-the-fly from a Chado database via a Drupal website.
Dynamism for references was easy to generate as the relevant JSONs were easy to understand (they had keys :-)).
Dynamism for tracks is driven by (unfortunately) the included Perl scripts (Drupal is a PHP language). For speed, I had to make a new Chado adaptor, a hack based on Scott’s DAS::Chado adaptor (it’s 10x faster but probably wouldn’t work for GBrowse).

It would be great therefore to know how JBrowse expect the track JSON to be present so that I could write PHP code to query Chado. That would be much faster than executing an external perl script and it would also a whole new dimension of being able to do some things...

Another really cool thing to know is how to implement redirects when clicking on a feature. I’m sure I’ve seen it somewhere...

thanks
a

by Alexie Papanicolaou

[jbrowsetix]

Documentation of ImageTrack JSON format

Thanks for your comments Joshua & Alexie. It’s been a long time coming, but I do intend to start documenting the JSON format very soon (e.g. this week). There is a huge caveat on this, which is that the JSON is liable to change in the next release and possibly future releases after that. Some might say it’s too early to document, given that the format is not stable, but we’re getting some eminently reasonable requests to do so, and I hate to turn developers away. So we will document the JSON format for release 1.1, with the understanding that it may not stay valid for very long.

Just for reference, this ticket actually refers to ImageTracks (server-rendered tracks used mostly for pixel-resolution histograms at the moment), whereas it seems like most of the comments are talking about FeatureTracks (client-rendered tracks for GFF/BED/etc types of data). To disambiguate, I’ve introduced another ticket that links back to this discussion:
http://jbrowse.lighthouseapp.com/projects/23792-jbrowse/tickets/66-documentation-of-featuretrack-json-format

by Ian Holmes

[jbrowsetix]

Documentation of ImageTrack JSON format

Answering one of my own questions:
click redirect:
"urlTemplate": "http://www.google.com/search?q={name}"

where i’ve seen {name} or {id} being used. from http://gmod.org/wiki/JBrowse_Tutorial#Load_Feature_Data

a

by Alexie Papanicolaou

[rbuels]

working draft of this is at http://gmod.org/wiki/User_talk:RobertBuels#Data_Format_Specification:_Fixed-Resolution_Tiled_Images, which will be moved to http://gmod.org/wiki/JBrowse_Advanced_Topics#Data_Format_Specification:_Fixed-Resolution_Tiled_Images in the next couple of days when JBrowse 1.3.0 is released.