Jamboree 18 July 2008-comments

From GO Wiki
Jump to navigation Jump to search

Comments on 1st Reference Genomes electronic jamboree


First I'd like to thank everyone who participated in the electronic jamboree last week, I think for a first attempt it was a great success! I think what was very helpful is that we all had just annotated the genes we discussed. To plan what to do next, I'd like to have any comments about the jamboree, in particular:

  • Time
  • Length
  • Opportunity to make comments (maybe we need a schedule with everyone given 5-10 minutes to comment or bring up an issue)
  • Did we (potentially;) resolve enough questions to replace a face-to-face meeting?
  • Should we redo this?
  • if yes: at what frequency? is monthly good
  • should we do this as part of the monthly conference call, all annotating specific a gene the week before? Or should we have a different time for this?


I think that the UK curators also found the annotation jamboree really beneficial, and any concern we had before the jamboree about having enough to discuss on the 2 genes was put to rest!

In the UK, the curators came and curated together, but had just the day to annotate the chosen genes before the call. The size of the group was ideal annotation discussions (there were about 8 of us), and was particularly helpful for curators who often annotate in isolation, where the potential for low-level annotation discussions is limited. However the general feeling was that in future jamborees it would help if we had all created our annotations a couple of days before the call, so that we had more time to review other group's annotation sets, and consider the issues to raise at the meeting.

I'd be very interested in doing this again, and agree that perhaps by having this type of meeting on a more regular basis, it may mean that annotation discussions would end up being more manageable/less stressful- instead of infrequent annotation meetings a year full of disagreements, we might be able to work to get many issues raised and resolved gradually?

We did have so much to discuss, that I agree that the calls have to be disciplined - it took one hour to discuss a gene on Friday, so 5-10 minutes to comment or bring up an issue sounds reasonable. If an issue needs much more time, then its likely that we need to create an action item and go away and find further information/examples.

For the human group, monthly jamborees might be a bit of a strain - in order for all reference genome species to have enough papers to participate, it looks as if the human group (and probably mouse/rat?) will often have a large pile of papers to annotate! How would quarterly (every three months) sound? This would mean that we would be sacrificing one reference genome conference call in three to just talk about a specific gene, but still have other calls to discuss other topics, or to chase up issues that took longer to resolve. The different UK curation groups could be free to meet together for these jamborees when possible.

As well as making detailed minutes (Rachael is currently working hard on the minutes for Friday's call), perhaps some of the issues raised could be collected into annotation pages on the GO wiki, with each issue timestamped - and resolutions either noted as 'issue agreed' or 'open' - with the idea that this could be slowly developed into an active, structured annotation document. We have many annotation discussions, where the decisions made are not always clear or universally agreed upon, along with pages of minutes; which are difficult to keep track of. Could we attempt to collect focused annotation discussions into a structured document, which can be quickly modified as we revisit an issue - although I realize that this would not be a trivial task.


I agree with Emily's point that the annotations need to be finished a few days before the jamboree so that we have time to look at the results prior to discussion. I think that doing this would go a long way towards keeping the discussion on track (and timely) as we would have more time to consider the points.

Quarterly jamborees sounds ok to me.


I thought the meeting went very well considering it was not face-to face. I think 2 genes was a good number, but that will depend on the genes that we are curating. I think quarterly would be a good time interval for the meetings, but I also think we do not want to do our annotations too far in advance. One of the things I found beneficial was that my annotations were still fresh in my mind.

Another thing that made this easier for me was that I had made all the annotations. That way I knew exactly what the rationale was for them when they were discussed. I don't know how other databases handled this, but perhaps we should have a 'champion' annotater for each gene from each MOD who would be the representative on these calls.


I think the meeting went very well and agree with all that have expressed having it on a quarterly basis, having the annotations made close to the meeting time and having only 2 genes.

I also think that this kind of exercise should be available and open to all curators that do GO annotations. We apply the standards for ref genome genes to all genes which we curate as part of various projects. Groups in various MODs could collectively curate the genes for the meeting and if not all members could attend (we could have) then they could take individual turns – something along the lines of what David was saying. With travel costs rising these meetings could be GO camps online. This too would argue for small numbers – 2 genes and spaced meetings – quarterly.

Finally and beyond annotations, I think it is of the essence to assure the quality of the vocabulary. I know there is a big effort to continuously improve upon the content and the depth of GO. Some of the issues raised during the meeting had more to do with the absence/presence of synonyms and the extent to which the definitions were or not clear – in other words, not very big things. Nonetheless, such meetings can bring up issues that might be easier to overlook within a bigger scheme of things and further help improve GO.


I agree, I thought that it was a very useful and productive meeting. I think quarterly is about the right period for these meetings.


I agree with the consensus so far, that a discussion every 3 months would be interesting and hopefully productive.

Reading Davids comments made me realise that actually it is difficult to feel engaged in discussions about a gene which you have not read any papers about. I think I am a bit confused about the 'champion' annotater idea. In the conference call it probably would be good if each group had discussed face-to-face what issues had arisen during the annotation and then one person aims to bring up each point. However, I think it will potentially mean that 1 person from each groups participates in the discussion and the other curators only half listen on the side lines. Especially if only 1 person has actually read the literature for the gene.

If we are doing this process more regularly then perhaps only one gene should be tackled for each call, if it is a well studied gene then possibly there will be enough papers for each curator in each group, even if there are 5-7 curators. Although this might lead to a rather over annotated gene, it would enable each curator to actively participate in the discussion. It would also provide a more effective way of improving annotation consistency across and within databases

Return to Electronic_jamboree_july18-200