How To Link To Gbrowse And Load Custom Track From External Data

Question

[Daniel Standage]

I am implementing a web frontend for a tool I've developed for annotating a particular type of genomic feature. The user can provide their own data via copy/paste or file upload, but I also have sequences for a few model organisms pre-loaded on the site. The tool generates a GFF3 file that is accessible via a web URL.

For users that provide their own data, they are of course on they own for visualizing and analyzing the results. But for the pre-loaded model organisms, I thought it would be useful to provide a links to public GBrowse browsers. For each data point, it would be great to provide a link that will take user to the relevant genome context view in GBrowse. In addition, it would be great if the links could somehow tell GBrowse where the data are and load them as a custom track for that user.

I've been playing with MSU's rice GBrowse instance, and I've come up with something that works. If I provide the following link (filled in with the appropriate values)...

http://rice.plantbiology.msu.edu/cgi-bin/gbrowse/rice/?start=$start;stop=$end;ref=$chr;eurl=$mygff3url

...then MSU's GBrowse will do precisely what I described--go to the relevant genome context view, populate a custom track with data from my site, and display the track. I don't know if this is the right way or best way to do it, but it seems to work.

I've had less success with the GBrowse instances at TAIR and MaizeGDB. I was unable to get any custom data to display at TAIR (even by manual upload), so they may have other issues. I was, however, able to get custom data to display at MaizeGDB by manual upload. Unfortunately, the little trick that worked for MSU does not work on MaizeGDB. It probably has to do with the fact that they are running different versions of GBrowse.

Is there any documentation for what I am trying to do? Is there a solution that will work across different versions of GBrowse?

Thanks!

Answer 1

[Scott Cain]

The rice and maize browsers are different versions of GBrowse (1 and 2 respectively), but they both support the eurl cgi parameter. For example, this url works for MaizeGDB:

http://gbrowse.maizegdb.org/gb2/gbrowse/maize_v2/?start=376000;stop=475999;ref=Chr1;width=800;eurl=http://5585rosemontway.com/sites/default/files/maize_chr1_section_0.gff

(I created a simple GFF3 file and hosted it on one of my webservers.) One trick is to identify the correct url to invoke. For example, the MaizeGDB instance doesn't have "cgi-bin" in it url, where many GBrowse instances do (it might work with cgi-bin/gb2 but I haven't tried it). The documentation for invoking public cgi GBrowse parameters is at:

http://gmod.org/wiki/GBrowse_Configuration/URL_schema

(which is not a url I knew off the top of my head; I searched gmod.org for "eurl".) Also helpful is using the "Bookmark this" functionality under GBrowse's File menu. It creates a URL that will exactly replicate the view you are looking at, which makes it easier to figure out how to generalize.

Answer 2

[Michael]

I have recently played around with GBrowse and found it fair enough to install my own instance. If you need full control over the appearance of genome tracks, it might be an option to install your own instance of GBrowse2 and aggregate some genome track for these organisms. This will add overhead in administration, computation and updates of data though, but provides you with a unified instance. It will depend on your requirements and capacities if that is a good solution.