Tripal User's Meeting 2017/02/03

Meeting Date: 
Friday, February 3, 2017
Attendees: 

Victor Unda, (WSU)

Chun-Huai Cheng (WSU)

Nic Herndon (UConn)

Michael Wynne (UConn)

Lacey-Anne Sanderson (Univ. of Sask)

Jobil Joseph (UConn)

Ethy Cannon (Iowa State)

Stephen Ficklin (WSU)

 
Agenda: 
  1. Thanks to those who participated at our PAG Meeting and Hackathon!

    1. 55-65 attendees at our Tripal Workshop at PAG

    2. 20 people at the Hackathon

    3. 30 people at the Tripal User's Meeting.

  2. Is anyone going to the Plant Biology 2017 annual meeting in Honolulu?  Would a Tripal User's meeting be good at this meeting?  Maybe a hackathon similar to PAG?

  3. Stress Testing

    1. Lacey is storing ~5 billion of records for natural diversity with fast response times but at the expense of data storage.

      1. Using materialized views and indexes to get this performance.

      2. At the expense of data storage (over 2TB to store).

      3. Is there a plan for storage expansion?

      4. Will that be a bottleneck?

      5. Option is to store materialized views in a non-SQL option. But will keep data in Chado as a "warehouse" for data integrity.

  4. Tripal v3 beta: planning for next week.  Feb 10th.

    1. Will have a migration/upgrade path available

    2. Will use drupal updates from here on out to support further updating.

    3. API for Tripal v3 will be stable but documentation for it is not yet ready.  

      1. Perhaps do a training meeting if interested on Tv3 API.

    4. What components have to be manually upgraded to Tripal v3 if customized? (Nic)... If you have a Tripal v2 site what do you need to manually adjusted.

      1. Initially there's nothing you manual upgrade.

      2. Switch nid_is to entity_ids in custom searches and materialized views

      3. Custom code that attaches custom panes/fields to the node.

      4. Move old Tv2 templates to GUI-based layouts.

  5. Tripal v2.1 stable -- delayed by need for testing.

 

    1. Currently working on Unit/Functional tests to help make releases go faster.

  1. Proposed changes to the Tripal Job Launcher (Ethy)

    1. Add TJL parameter ‘max_jobs’ and function tripal_max_jobs_exceeded().

      1. This is meant as protection when turning on parallel mode to ensure that the system is not overrun with jobs

    2. Add TJL parameter ‘single’ to mean only one queued job should be executed each time the TJL executes.

--parallel=1 --single=1  ensures that each job will be executed in parallel.

    1. Add timestamps to messages.

    2. ACTION ITEM: Ethy to submit pull request :-)

  1. Report on GO loading problems (Ethy)

    1. Error is due to a cyclical cvterm_relationship which is not in the GO obo file.

      1. Didn’t seem to be in the original file?

    2. Guessing this is left over from loading an older GO obo file; cvterm_relationships are not cleared when a new version is loaded and the new obo file added a relationship that caused the cycle.

      1. Stephen has had this happen on a fresh install

    3. Could this be related to this? https://github.com/erasche/chado-schema-builder/pull/3 (Anthony Bretaudeau)

    4. ACTION ITEM: Stephen to look into this

    5. ACTION ITEM: Ethy to send the particular terms that she suspects

  2. Google Summer of Code 2017

    1. Email from Robin Haw states:

      1. Our team name since 2011 has been Open Genome Informatics. Since 2011, the Open Genome Informatics team has served as an "umbrella organization" to a variety of open bioinformatics projects, including GMOD and its software projects -- GBrowse, JBrowse, etc.; Galaxy; PortEco; Reactome; SeqWare; WormBase; DockStore and others.

    2. http://gmod.org/wiki/GSoC

    3. If you have a project idea for which you would like to mentor a student, please contact Robin Haw, cc’ing Scott Cain and Marc Gillespie

      1. If you don't have their email's Stephen can forward them to you.. just ask.

      2. Just don’t want to paste them here since our notes go online

    4. Application Due Feb 9th!... email Robin this week with a project idea.

    5. Stephen participated last year and had a very good experience with a student from India.

      1. https://summerofcode.withgoogle.com/archive/2016/projects/5902985066446848/

      2. Mailing List

    6. Get rid of gmod-tripal, gmod-tripal-devel.  Use only gmod-tripal-announce to make meeting, outreach and new version announcements.

    7. There was concern that “what do you think” questions might feel out of place on the issue queue on github

      1. Suggestion was to use biostars

      2. Also suggestion to use custom issue tags to make it obvious we want these questions on github to make it more inclusive

  3. Upcoming Training Topics

    1. Using the Tripal API

    2. How to write your own module.  (An interest for Nic/TreeGenes)

    3. Using the bulk loader

    4. How to make your custom modules generic enough for sharing with others.

      1. There are some notes on this in the PAG Hackathon Notes.

      2. ACTION ITEM: Lacey to make this into a full document

    5. Where does Drupal end and Tripal begin?

    6. Using Views to create custom searches.

    7. How others have created custom search pages.  (Ethy -- late March is better.)

    8. Other ideas?

    9. ACTION ITEM: Stephen send out doodle poll for training.

  4. Round table:  Anyone with new modules, a new site, anyone needing help with any particular topic?

    1. Maybe a CyVerse question - can data uploaded to TACC with Tripal Workflows (using Galaxy API), be accessed without Tripal (e.g., using Agave API), and vice versa? (Nic)

      1. ACTION ITEM:  Find out how much interaction there is between the Galaxy and Agave API.

      2. Currently an option is to store your files at TACC and then tell Tripal Workflows that is where it is.

      3. Stephen recommends storing data physically close to the Galaxy instance that you intend your users to use with Tripal Workflows.

        1. For example, if using the public galaxy then store your data at CyVerse

      4. If you are looking at having a public dataset available separately from your Tripal site, you may want to talk to the Legume Federation (Ethy) to see how they are doing it

        1. They’ve put a lot of thought into how to store the metadata associated with a file

  5. Reminders

    1. Anyone can put Tripal related repositories on the Tripal Github organization

      1. We would love to add you to this organization to make it easier to see what modules are available and what progress is occurring on various modules

      2. If you are listed as private, please consider making yourself public to show amazing our community is and how many active members we have

      3. We can organize people into teams to ensure that your module is still attributed to your group if you put it under the Tripal organization

    2. Please remember to add your modules to the tripal.info site (let us know if you need an account).

      1. There is page in every Tripal site that pulls from this site and shows admin what extension modules are available

      2. To make your module visible, please add it to this site

    3. Let us know if you want your site featured on the Tripal.info site.

      1. Only sites we have specific permission to highlight will be shown on here