Uploaded image for project: 'IGB'
  1. IGB
  2. IGBF-3007

Implement next iteration of track hub converter interface

    Details

    • Type: New Feature
    • Status: Closed (View Workflow)
    • Priority: Major
    • Resolution: Done
    • Affects Version/s: None
    • Fix Version/s: None
    • Labels:
      None
    • Story Points:
      2
    • Sprint:
      Fall 7 2021 Nov 8 - Nov 24, Fall 8 2021 Nov 29 - Dec 10, Fall 9 2021 Dec 13 - Dec 24

      Description

      Implement design of next iteration of track hub table interface.

        Attachments

        1. igb_icon.png
          igb_icon.png
          17 kB
        2. Logo_roundedEdge.psd
          593 kB
        3. Logo.svg
          145 kB

          Issue Links

            Activity

            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment -

            [~aloraine], I was wondering if we might use a different icon than the default IGB icon I used in the design mockup. It looks a little off from the site theme color-wise, and letters in the icon are difficult to make out, since the icons are small to accommodate the current genome version line spacing. We could increase the line spacing to allow for a bigger icon size, but it would be nice if there were an SVG file of the icon somewhere that I could edit to change the colors a bit. Do you know if one exists anywhere?

            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - [~aloraine] , I was wondering if we might use a different icon than the default IGB icon I used in the design mockup. It looks a little off from the site theme color-wise, and letters in the icon are difficult to make out, since the icons are small to accommodate the current genome version line spacing. We could increase the line spacing to allow for a bigger icon size, but it would be nice if there were an SVG file of the icon somewhere that I could edit to change the colors a bit. Do you know if one exists anywhere?
            Hide
            ann.loraine Ann Loraine added a comment -

            I don't know of a different image.

            Show
            ann.loraine Ann Loraine added a comment - I don't know of a different image.
            Hide
            ann.loraine Ann Loraine added a comment -

            Maybe also ask Nowlan Freese.

            Show
            ann.loraine Ann Loraine added a comment - Maybe also ask Nowlan Freese .
            Hide
            nfreese Nowlan Freese added a comment -

            Philip Badzuh - there are a number of different versions of the IGB logo in the Dropbox > IGB > Logo > IGB folder. I attached an example svg and psd that were in the folder.

            Show
            nfreese Nowlan Freese added a comment - Philip Badzuh - there are a number of different versions of the IGB logo in the Dropbox > IGB > Logo > IGB folder. I attached an example svg and psd that were in the folder.
            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment -

            Using the svg file that Dr. Freese provided as a starting point, I have designed an icon that can be used to open genomes in IGB. I have attached it to this ticket as igb_icon.png. I think the white background and avoiding lettering in a small image size makes this a better option. Please let me know your thoughts.

            Regarding the integration of the new endpoint to determine whether or not UCSC genome versions listed in the table have IGB equivalents and can, therefore, be opened in IGB, I have been facing an issue with correct mapping. After testing the facade endpoint, this appears to be an issue on the frontend side, so I'm working to debug my javascript.

            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - Using the svg file that Dr. Freese provided as a starting point, I have designed an icon that can be used to open genomes in IGB. I have attached it to this ticket as igb_icon.png. I think the white background and avoiding lettering in a small image size makes this a better option. Please let me know your thoughts. Regarding the integration of the new endpoint to determine whether or not UCSC genome versions listed in the table have IGB equivalents and can, therefore, be opened in IGB, I have been facing an issue with correct mapping. After testing the facade endpoint, this appears to be an issue on the frontend side, so I'm working to debug my javascript.
            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment - - edited

            Progress update:

            • I have fixed the bug related to incorrect determination of whether or not genomes in the table can be opened in IGB. Some metrics to note:
              • Of all the available genomes in the table, 292 can be opened in IGB, while 1655 cannot.

            TODO:

            • Implement opening genomes in IGB
            • Move search bar to top and center
            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - - edited Progress update: I have fixed the bug related to incorrect determination of whether or not genomes in the table can be opened in IGB. Some metrics to note: Of all the available genomes in the table, 292 can be opened in IGB, while 1655 cannot. TODO: Implement opening genomes in IGB Move search bar to top and center
            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment -

            Please see my changes here.

            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - Please see my changes here .
            Hide
            ann.loraine Ann Loraine added a comment -

            Small change request:

            Please change file name "igb_logo.png" to "open_in_genome_browser.png" (or other descriptive name) to avoid confusion as to how this file is being used.

            Once file is renamed and other files are changed to use the new name, please submit PR.

            Everything looks fine - very low-risk changes at this point!

            Show
            ann.loraine Ann Loraine added a comment - Small change request: Please change file name "igb_logo.png" to "open_in_genome_browser.png" (or other descriptive name) to avoid confusion as to how this file is being used. Once file is renamed and other files are changed to use the new name, please submit PR. Everything looks fine - very low-risk changes at this point!
            Hide
            ann.loraine Ann Loraine added a comment -

            Moving back to "to-do" to make these (hopefully lightweight) changes.

            Show
            ann.loraine Ann Loraine added a comment - Moving back to "to-do" to make these (hopefully lightweight) changes.
            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment -

            Please see PR here.

            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - Please see PR here .
            Hide
            ann.loraine Ann Loraine added a comment -

            PR is merged. New code is deployed to BioViz main for testing.
            attn: Philip Badzuh and Omkar Marne

            Show
            ann.loraine Ann Loraine added a comment - PR is merged. New code is deployed to BioViz main for testing. attn: Philip Badzuh and Omkar Marne
            Hide
            ann.loraine Ann Loraine added a comment -

            Please provide testing instructions.

            Show
            ann.loraine Ann Loraine added a comment - Please provide testing instructions.
            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment -

            Please see PR here.

            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - Please see PR here .
            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment -

            To test, once this has been merged and deployed to bioviz.org, visit https://www.bioviz.org/public-trackhubs.html and ensure the following:

            • The table loads.
              --You will need to 'clear site data' from developer tools -> application -> storage
              --Note that this may take several minutes, as some of the public hub servers are slow to respond.
            • The search filters the table in real time.
            • Clicking on the Q copies the converted/quickload data source representation of the trackhub.
            • Adding that copied URL as a data source in IGB works.
            • Clicking on the genome icon when IGB is running (for hubs that have been added as data sources in IGB), results in the genome opening in IGB.

            Please note any bugs, suggestions, or other feedback. Thanks in advance.

            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - To test, once this has been merged and deployed to bioviz.org, visit https://www.bioviz.org/public-trackhubs.html and ensure the following: The table loads. --You will need to 'clear site data' from developer tools -> application -> storage --Note that this may take several minutes, as some of the public hub servers are slow to respond. The search filters the table in real time. Clicking on the Q copies the converted/quickload data source representation of the trackhub. Adding that copied URL as a data source in IGB works. Clicking on the genome icon when IGB is running (for hubs that have been added as data sources in IGB), results in the genome opening in IGB. Please note any bugs, suggestions, or other feedback. Thanks in advance.
            Hide
            ann.loraine Ann Loraine added a comment - - edited

            Note: updated playbook line that configures back end domain address. This may not have been strictly necessary, but did it to highlight how deployment depends on modifying hub-related js code following checkout.

            See this commit

            Show
            ann.loraine Ann Loraine added a comment - - edited Note: updated playbook line that configures back end domain address. This may not have been strictly necessary, but did it to highlight how deployment depends on modifying hub-related js code following checkout. See this commit
            Hide
            ann.loraine Ann Loraine added a comment -

            Deployed new code to bioviz main. Ready for testing as per protocol described above.

            Show
            ann.loraine Ann Loraine added a comment - Deployed new code to bioviz main. Ready for testing as per protocol described above.
            Hide
            ann.loraine Ann Loraine added a comment -

            Attn Philip Badzuh: The new code is deployed on bioviz main and is using translate.bioviz.org as the back end:

            www.bioviz.org ec2-user $ more util.js 
            export const BACKEND_DOMAIN = "translate.bioviz.org";
            

            I visited the table page in Safari, and I saw a message appear indicating that the page was attempting to get data from UCSC. The spinner appeared and kept going, but eventually quit and the page did not update.

            Show
            ann.loraine Ann Loraine added a comment - Attn Philip Badzuh : The new code is deployed on bioviz main and is using translate.bioviz.org as the back end: www.bioviz.org ec2-user $ more util.js export const BACKEND_DOMAIN = "translate.bioviz.org" ; I visited the table page in Safari, and I saw a message appear indicating that the page was attempting to get data from UCSC. The spinner appeared and kept going, but eventually quit and the page did not update.
            Hide
            ann.loraine Ann Loraine added a comment -

            Is there a faster endpoint you could hit to create the table and its rows? Within the row cells, you could possibly invoke some jquery-type code that would update cell contents when the slower-to-return endpoints come back with some data?

            The track hub table at UCSC is probably showing static data, or data retrieved from a database. Their design can tolerate slow endpoints. Maybe do some google searching to find out how table designers handled slow data retrieval?

            Show
            ann.loraine Ann Loraine added a comment - Is there a faster endpoint you could hit to create the table and its rows? Within the row cells, you could possibly invoke some jquery-type code that would update cell contents when the slower-to-return endpoints come back with some data? The track hub table at UCSC is probably showing static data, or data retrieved from a database. Their design can tolerate slow endpoints. Maybe do some google searching to find out how table designers handled slow data retrieval?
            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment -

            The bottlneck isn't the UCSC API, but rather the hub.txt on which it depends, which is hosted on user servers that we have no control over. The culprit request is to https://api.genome.ucsc.edu/list/hubGenomes?hubUrl=http://www.mircode.org/ucscHub/hub.txt, and the problem is that mircode.org is very slow/keeps timing out.

            Some options:

            • Set a timeout on requests for public hub data and ignore ones that are too slow. It's likely that their data would be slow to load in IGB as well, which would only lead to user frustration.
            • Set the shorter timeout to get most of the public hubs loaded, and then reattempt slow sites asynchronously
            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - The bottlneck isn't the UCSC API, but rather the hub.txt on which it depends, which is hosted on user servers that we have no control over. The culprit request is to https://api.genome.ucsc.edu/list/hubGenomes?hubUrl=http://www.mircode.org/ucscHub/hub.txt , and the problem is that mircode.org is very slow/keeps timing out. Some options: Set a timeout on requests for public hub data and ignore ones that are too slow. It's likely that their data would be slow to load in IGB as well, which would only lead to user frustration. Set the shorter timeout to get most of the public hubs loaded, and then reattempt slow sites asynchronously
            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment -

            Regarding the issue in safari, I just checked my changes, and it seems that I changed the domain variable for the wrong endpoint. I will work to fix this now.

            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - Regarding the issue in safari, I just checked my changes, and it seems that I changed the domain variable for the wrong endpoint. I will work to fix this now.
            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment -

            Please see the PR to fix the issue here.

            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - Please see the PR to fix the issue here .
            Hide
            ann.loraine Ann Loraine added a comment -

            Merged and redeployed to bioviz main for testing.

            Show
            ann.loraine Ann Loraine added a comment - Merged and redeployed to bioviz main for testing.
            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment -

            I will need to make an additional change to the bioviz code to ensure compatibility with the new changes made to the hub facade api in IGBF-3010. Moving back to to-do.

            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - I will need to make an additional change to the bioviz code to ensure compatibility with the new changes made to the hub facade api in IGBF-3010 . Moving back to to-do.
            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment -

            Please see the required changes here.

            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - Please see the required changes here .
            Hide
            ann.loraine Ann Loraine added a comment -

            Changes look good. Please submit PR.
            attn: Philip Badzuh

            Show
            ann.loraine Ann Loraine added a comment - Changes look good. Please submit PR. attn: Philip Badzuh
            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment -

            Please see PR here.

            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - Please see PR here .
            Hide
            ann.loraine Ann Loraine added a comment -

            PR is merged and new code is deployed to bioviz.org. Ready for testing.

            Show
            ann.loraine Ann Loraine added a comment - PR is merged and new code is deployed to bioviz.org. Ready for testing.
            Hide
            omarne Omkar Marne (Inactive) added a comment -

            After visiting https://www.bioviz.org/public-trackhubs.html the table didn't load.

            Philip Badzuh Please check.

            Show
            omarne Omkar Marne (Inactive) added a comment - After visiting https://www.bioviz.org/public-trackhubs.html the table didn't load. Philip Badzuh Please check.
            Hide
            ann.loraine Ann Loraine added a comment -

            Omkar Marne - move back to To-Do please.

            Show
            ann.loraine Ann Loraine added a comment - Omkar Marne - move back to To-Do please.
            Hide
            pbadzuh Philip Badzuh (Inactive) added a comment - - edited

            Based on my investigation of the problem observed by Omkar, the cause seems to be related to the hub facade's Django CORS settings. I have addressed this issue in IGBF-3036. Moving this ticket to needs testing.

            Show
            pbadzuh Philip Badzuh (Inactive) added a comment - - edited Based on my investigation of the problem observed by Omkar, the cause seems to be related to the hub facade's Django CORS settings. I have addressed this issue in IGBF-3036 . Moving this ticket to needs testing.
            Hide
            omarne Omkar Marne (Inactive) added a comment - - edited

            https://www.bioviz.org/public-trackhubs.html is accesible and trackhub table has loaded.

            Closing the ticket.

            Show
            omarne Omkar Marne (Inactive) added a comment - - edited https://www.bioviz.org/public-trackhubs.html is accesible and trackhub table has loaded. Closing the ticket.

              People

              • Assignee:
                omarne Omkar Marne (Inactive)
                Reporter:
                ann.loraine Ann Loraine
              • Votes:
                0 Vote for this issue
                Watchers:
                4 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: