Minutes from JBrowse meeting, 1/9/12

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

Minutes from JBrowse meeting, 1/9/12

Robert Buels
Minutes mostly by Nomi Harris, with some edits and expansions by Rob.


Rob, Scott, Nomi, Suzi, Lincoln, Chris C., Ed


Rob's official start day is probably tomorrow (waiting for official
Working on an integration test suite in Selenium (had to do it in
Python--not quite as nice as the Perl testing stuff). (This includes
some UI testing--Selenium lets you do this by basically launching a
browser that you can remote-control.)
First will integrate Natalie Fox's faceted track selected
Then will work on merging in Mitch's JSON refactoring stuff. Hopes to
have that done before PAG.

There's an intern named Julie Moon working with Lincoln on documentation
of URL API (for purposes of embedding) and will work out a mechanism for
turning off the JBrowse controls, so that people will be able to embed
static JBrowse shots in web pages that let the user move around only in
very limited ways.

Rob is going to PAG and was planning to fly home the morning of the
19th. It may be hard for him to arrange to stay longer to join us for
part of the WebApollo Hackathon, but he will see what he can do.

Some discussion about folding, packing, and overlay of tracks and
subtracks.  The overlay option makes features no longer bump
eachother, but overlay eachother.  This is especially useful for
multiple wiggle tracks.

Suzi asked: Lincoln mentioned about moving tracks to overlay each other,
and also how it would be nice to have the ability to do set operations
(union, etc.) on tracks. Where is this on your priority list?
Rob said it's on his radar but he's not sure what the priority is. We
should talk to Ian about that.
Suzi: What does the UCSC browser offer in terms of set operations?
Lincoln: UCSC has various "packing" operations (full/squished/compact)
that control how collision control works. There's a new packing
operation called overlay (or something like that) with a transparency
value, so that instead of bumping each other, features are superimposed
on each other with transparency. This is useful for wiggle tracks--can
put them into a sort of merged profile.
Suzi: Is it dynamic?
Lincoln: No--not integrated with drag.
Lincoln: It would be great if you could dump tracks on top of each
other, and if you did that, it could pop up a menu to let you do set
Suzi: Yes, and you might want to create a new track and put your new
composed track into it.
We will talk to Gregg about this (he implemented something like this in

Ed: Tracks are listed in order they appear in the configuration file. He
added something to allow you to sort the track order.
Suzi: Would like to revisit different ways to index tracks. Don't like
the labels. Would like to have track information/controls aligned with
the tracks.
Ed: We once had a discussion about having track labels float to the left
so they don't use up real estate.

Suzi: Would the work Gregg did on making features show the individual
bases be useful in JBrowse?
(Currently in JBRowse, the DNA track shows up if you zoom in enough, but
you don't see bases in other tracks.)
Ed: Well, this is most useful when editing, but could also be nice if
you saw bases when you zoomed in enough. Right now, though, it only
shows DNA sequence when you select a feature, so if we wanted to show
bases in JBrowse, would have to either (1) add WebApollo's selection
capability to JBrowse (which might be useful anyway), or (2) show bases
in all tracks once you zoom in enough.
Rob: It would make sense to add both selection and more basepair-showing
capability to core JBrowse.

Hackathon objective: add ability to get data from Chado into
JBrowse/WebApollo. This will greatly increase the number of datasources
that JBrowse/WebApollo will have access to. (Can already read UCSC and