Details
-
Type:
Task
-
Status: Closed (View Workflow)
-
Priority:
Minor
-
Resolution: Done
-
Affects Version/s: None
-
Fix Version/s: 10.1.0
-
Labels:None
-
Story Points:1
-
Epic Link:
-
Sprint:Spring 4, Spring 9, Spring 10
Description
Situation - While testing IGBF-3528 we discovered that when the same UCSC REST or UCSC DAS track (for example, refGene) was loaded in two different genomes, the track would disappear.
Task: Investigate this issue and propose a fix if possible.
Note: This issue appears in 9.1.10 when loading UCSC DAS data as well as with the recent implementation of UCSC REST.
To reproduce the issue:
- Select the mouse genome.
- Select Available Data > UCSC (DAS) > refGene
- Click Load Data
- Change to the human genome.
- Select Available Data > UCSC (DAS) > refGene
- Click Load Data
- Go back to the mouse genome
- The refGene added from UCSC should be missing from the Data Management Table. The data should still be present as a track, but grayed out as if it had not been loaded.
Attachments
Issue Links
- relates to
-
IGBF-3528 Implement logic to load data for genePred file type
-
- Closed
-
Activity
PR is merged. Installers are built and deployed to bioviz.org "early access" section.
Pull request is raised: https://bitbucket.org/lorainelab/integrated-genome-browser/pull-requests/1012. Please review and let me know if there are any issues.
Tested on Mac with Sravani's branch.
Both of the issues described in the description and Sravani's comment are fixed. Tracks are no longer greyed out when loading the same data set (UCSC REST) between two different genomes. If using DAS and UCSC REST the same datasets (for example, refGene) they no longer disappear.
I'm unsure of why the getAnnotStyle in DataSet.java was originally being passed "method" twice. We had previously had some tickets looking into issues with how track names were being set (see IGBF-1933), but the code in Sravani's branch has not been modified since David worked on it. I tested BioViz Connect data, online Quickload data, local Quickload data, and local files and all of the track names appeared as I expected.
Ready for pull request.
There are two issues mentioned in this ticket, I have investigated the first one. Steps to reproduce first issue are:
- Select any genome that has both UCSC DAS and UCSC REST Tracks available.
- Select the same track from both the data providers.
- Before implementing this ticket https://jira.bioviz.org/browse/IGBF-3703 while loading the empty tiers itself, only one track was visible, the other one would disappear as the tracks had the same names. However, after the implementation of that ticket, this issue was resolved as the track name was changed for the REST tracks.
- Clcik on the Load data button, now you will see one track has disappeared as the track name is being rewritten by some other method. This issue is fixed now. Updated code is available at branch: https://bitbucket.org/jaya-sravani/integrated-genome-browser/branch/IGBF-3584. Tested and working as expected.
Screeshots before the fix:
Empty Tier Load
After load data
Screeshots after the fix:
Empty Tier Load
After load data
Tested using main branch installer on Mac.
Loaded refGene track from UCSC REST and from DAS in the mouse genome. Neither track disappeared (correct behavior).
Loaded refGene track from UCSC REST in mouse and then switched to human and loaded refGene from UCSC REST. When I returned to the mouse genome the track was not greyed out or removed (correct behavior).
Closing ticket.