13 May 2014 PAINT Conference Call

From GO Wiki
Jump to: navigation, search

May 13 2014 PAINT conference call

Participants: Rama, Donghui, Paul T, Moni, Judy, Kimberley, Suzi, Huaiyu, Pascale, Karen, Chris, Li

Upcoming “PAINT Jamboree”

  • Paul gave a brief introduction about the Jamboree - More intense PAINT curation before the Barcelona meeting. Ball park resource is 30 people week time between June 1 to early Oct. Question is whether to have a kickoff meeting to have everyone together or do it through skype or google hangout.
  • Rama and Judy - More useful and efficient to have everyone together in the same room. Therefore a kickoff meeting will be good.
  • Rama: Since most of us are in California, we should have the meeting here.
  • Huaiyu: Date? Will set up a doodle poll to decide. Participants? The same group on the PAINT call.

Proposed new functionality?: Adding or removing experimental annotations real-time in PAINT

  • Paul: During PAINT annotation, sometimes missing an annotation or an incorrect annotation can prevent curators to propagate the terms correctly. They have to go through the term challenge or other communications to get the annotation added or corrected. This usually takes weeks or even months before they can come back to the family and reannotate it. More efficient way is to do real time "literature annotation" so that these terms can be propagated right away. The question is whether this is a show-stopper without this functionality, and whether we would like to have it for this jamboree.
  • Suzi: The question is whether we should have it now. Since we have survived without it so far, can we wait for another 6 months when the javascript paint is ready.
  • Huaiyu: For the jamboree, is it helpful to test the paint software thoroughly so that there is no surprise. Software test is a bit different from curation, because you pay attention to different things.
  • Suzi: Should not spend too much time on testing if it will delay the jamboree.
  • Rama mentioned about taxon constraint problem. Suzi said the paint software worked fine but the problem was due to the incompleteness of the taxon constraints.
  • Karen: Adding new functionalities is helpful but it depends on how big the task is. It is more practical to use the jamboree to figure out what new functionalities to add.
  • Suzi: Agree with Karen. Would like to use jamboree to collect information to guide the development of javascript Paint.
  • Paul: Two options to deal with this issue. Option 1 is current practice to go through the challenge and wait for the literature curation to add or remove the annotations. The cycle can be months, and then come back reannotate the family. Option 2 is to add or remove in paint and propagate right away. Option 1 is OK. It gives you the correct paint annotation but not complete, while option 2 will allow you to provide more complete paint curation.
  • Judy: There is an option 3 as discussed in the (off-site) meeting, that is to coordinate between the paint curation and literature curation, so that the literature curation will precurate the families before paint curation.
  • Huaiyu: The combination of Option 2 and 3 will be ideal. Right now, the cycle is too long. Often times the challenge is not responded for months.
  • Karen: Used to curate transcription families that had been curated by me in literature curation, so there is usually no need to add more annotations. These days, I am curating cilia families. Not all annotations are in mouse. It will be helpful to be able to add annotations in PAINT. But don't think it is urgent.
  • Huaiyu: It happens probably every 3-4 families. It is possible to use challenge to remove incorrect annotations that prevent me from propagation. It is difficult to add a new annotation.
  • Pascale: It happens at about the same rate. The problem is that after a month or two, it takes a lot of time to recurate the family, especially when I start to ramp up and curate a lot of families.
  • Li: Agree with Pascale that it is very inefficient to recurate after a month or two.
  • Paul: To summarize, we do encounter this problem. It is a nice feature to have, but not a show-stopper. We should look into how much time will be required to add the feature, and decide.