P. Nextgen collaboration

Meeting 18/11/2015

Participants: Naama Menda, Lukas Mueller, Marie-Angélique Laporte, Léo Valette

1. Workflow for updating ontologies 

Naama: We would like to have a fast way of assigning IDs to new traits and variables since CassavaBase users require constantly adding of new terms. We do not expect to be adding large number of traits in the near future, but we have to make sure we don't create conflicts. The options we see are either CO is responsible for giving us new IDs upon request, or each developer (right now it's just BTI and Fola, but may change in the future) gets a range of IDs to use.

Leo: CO will need a way of tracking the assigned IDs and making sure others are following it. What if different developers create the same trait with 2 different IDs? 

Naama: will have to create new issue for each new term, maybe using the github issue tracker at the Planteome repo, then it is up to each developer to check the tracker for new traits, and up to each one to add a new issue when introducing a new term.

Marie: What about the ID ranges for traits/variables, methods, scales

Naama: It's nice to have, but not that important now that we have 4 different namespaces

Lukas: both options are good for us, we can work with whatever is better for Crop Ontology.

Leo and Marie will discuss this and will get back with a preferred method . This is important since we will work in the near future on editing many more ontologies, currently we are looking at banana, yam , and sweet potato.


2. Working version of the ontology

CO works with the trait dictionary, since this is an IBP requirement. The problem is that when converting from the excel template to OBO things get lost sometimes. Another problem is that it is hard to track changes this way.

 Leo and Marie mentioned that they have much more information in the TD than in the OBO. Need to look into adding fields in the OBO that can be parsed into constructing a TD file with all the data that is normally not stored in Chado or part of the OBO fields, for example create a string with certain delimiters that would go into a 'comment' OBO field.

Lukas: we need to have one working copy that is edited. 

Leo: we have problem with version control of the trait dictionary, need to track changes made by users

Lukas: how about use the CO web site for this? It will be the central repository. the web site looks great and has very good visibility, people are already used to using it. If you can create a copy of the web site and set it up as a test site with htaccess password. Also look into making nightly backups to allow reverting if needed. BTI can possible host this test site.

Leo and Marie will discuss if the CO website can be the central repository and the implications in terms of the structure and the content of the input file (OBO/Excel) and in terms of conversion routines (adaptation of the TD2OBO routine, coding of a OBO2TD routine).

3. Continuing collaboration

Lukas offered to invite here Leo and Marie. they will be in Corvalis OR next month, they will check with Elizabeth about changing the itinerary to make a stop for a few days in Ithaca NY. Marie suggested the week of November 16th.