question? trackhubs 'forgetting' their selected subtracks upon modifying another track

3 views
Skip to first unread message

Jules Kerssemakers

unread,
Jun 30, 2015, 11:58:44 AM6/30/15
to gen...@soe.ucsc.edu, Vladimir Kuryshev
Dear UCSC'ers,

I have a question / minor annoyance where I'd like to ask for your community wisdom how to get around it.

We use the new Genome-Browser-in-a-Box (GBiB) locally to work with patient data. The patient data itself we have divided into trackhubs per data-type (* layout below), with separate tracks per trackhub for different sub-population (tumour, normal, etc). Each patient then has a subtrack inside these population-tracks.

The reason our tracks are layed-out like this is that we first explore specific loci:
"is anything going on at geneX in datatype Y?"
if so, we want to then 'zoom in' on those patients, and add additional datatypes for those patients that have something.
The annoying thing is that each time we add a subtrack from a second datatype, the first datatype is hidden again, so we can't compare them side-by-side.

In the public data-sets, we're used to being able to selectively disable/enable specific subtracks, e.g. Genes and Gene Predictions > GENCODE > Gencode v19 > Comprehensive set > submit
If I then go to another public track (e.g. Expression > Encode RNA-seq tracks > CSHL Long RNA-seq > (select some cell lines) > submit, the Genome-Browser remembers I was also viewing the Gencode V19 comprehensive subtrack, so it displays both the gencode subtrack, and the cell line subtracks I just selected.
We have tested this behaviour on both the public (euro) mirror, and our local GBiB, and this behaviour is consistent (as we expected)

We've done some tryouts to try and nail down the exact behaviour of trackhubs:
a) If we select a (single-patient) subtrack for viewing from one data-type-trackhub, and then try to load that same patient's subtrack for a different datatype (i.e. also from a different trackhub), the original selection is 'forgotten'.
Instead of showing both subtracks from the different trackhubs, the first subtrack's entire track resets to default ('hide' in our trackhub), and only the second subtrack is displayed.

b) Also if I select the second subtrack from the same Trackhub, but a different track, the same happens: the first subtrack is forgotten and the entire first track resets to 'hide'.

c) Even within the same track, if I click down to the same subtrack-details again, my original selection is forgotten, and all the subtracks are selected, instead of just the one I had selected earlier.
(Although this seems to be consistent with the built-in tracks: those also don't remember which subtrack I had selected, just which tracks.)

d) with a, b and c, the built-in tracks I had selected remember their selection (e.g. GenCode v19 comprehensive track remains selected throughout).

Ideas we had so far (please add to them :-) ):
- Merging all three Trackhubs into one trackhub won't change anything. As we described: even selecting subtracks from two different tracks within the same trackhub overrides the first selection.
- Having one merged trackhub, but organised as one track per patient, with subtracks per datatype. Would work for the per-patient view, but then we won't have the ability to show "all mRNA-seq data", because we can't select all the mRNA-subtracks across the patient tracks.
We're a bit stumped how to have our cake and eat it too, so please advise! :-)



To me it seems that in trackhubs, the slected-subtrack-selection information is overridden with each submit, whereas in the public/built-in tracks the subtrack selection is updated/appended.
1) Is there a reason for this difference?
2) More specifically: is there a way to change this so that I can select multiple subtracks from different tracks?
3) if 2 isn't possible, how should I organise my trackhub to select multiple patients _and_ multiple datatypes simultaneously?



* Trackhub layout:
For hopefully obvious reasons I cannot share the trackhubs themselves (patient confidentiality etc.), but I think I can share our layout:
- Trackhub 1: 450K data
-- Track 1a: normal samples
--- subtracks-per-patient
-- Track 1b: tumour samples
--- subtracks-per-patient

- Trackhub 2: MCIp data
-- Track 2a: normal samples
--- subtracks-per-patient
-- Track 2b: tumour samples
--- subtracks-per-patient

- Trackhub 3: mRNA-seq
-- Track 3a: normal samples
--- subtracks-per-patient
-- Track 3b: tumour samples
--- subtracks-per-patient


Reply all
Reply to author
Forward
0 new messages