NetworkMeeting7

Network Meeting 7: Building an Ontology with Collaborative Protege

Contents:

1. Meeting Details
   1. Dates/Time
   2. Meeting Location
   3. Accommodation (in Cambridge)
   4. Costs
2. Programme of speakers and presentations
3. Attendees
4. Outcome
5. References
6. Organisers

1. Meeting Details

In this meeting we will, as a group, take part in an exercise to actually build some ontology in a collaborative fashion. The ostensible objective is to build a portion of the devices branch of the Ontology of biomedical Investigations (OBI, https://wiki.cbil.upenn.edu/obiwiki/index.php?title=HomePage). In addition, we wish to examine the process of collaborative ontology building itself. This is a continuation of our themes of exploring the process of ontology building and in particular collaborative ontology building. In NetworkMeeting5 a small group of us performed a normalisation of OBO's cell type ontology "on paper". This time we'll do it for real. The interactions we make with each other and the usage we make of the Collaborative Protege tool ( http://protege.stanford.edu/doc/collab-protege/#UsersGuide) will be recorded for later analysis. In addition, we might deliver some useful device ontology at the end.

We've chosen the instrument and devices section within OBI as all the scientists in OntoGenesis will use a device of some sort. The non-life scientists in the group would also have something to say about such an ontology. It is rare for an ontology to exist that affords such a wide opportunity to a group of scientists.

OBI is a collaboratively developed owl-DL ontology to support the ontological annotation of biological experiments. It covers a wide range of terms from multiple omics and biomedical domains. Specificly, the ontology models experimental design, protocols, instrumentation, biological material, data and types of analysis performed on data. Developed under the OBO Foundry initiative, OBI endorses the foundry principles ( http://www.obofoundry.org/wiki/index.php/OBO_Foundry_Principles) and makes use of the BFO top level ontology. In OBI, special working groups (branches) tackle the following subtrees concurrently:

Independent continuants:

• biomaterial

  instrument and device

  digital entity

  plan

Dependent continuants:

• function

  quality

  roles

Occurrents:

• planned process

  data transformation

To ensure consistent work across branches OBI developers agreed on common class naming conventions ( http://obofoundry.org/wiki/index.php/Naming), a minimal set of administrative and semantic metadata ( https://wiki.cbil.upenn.edu/obiwiki/index.php/MinimalMetadata) and methods for cross referencing other OBO foundry ontologies ( http://obi-ontology.org/page/MIREOT).

Workflow:

The envisioned flow path for the seventh Ontogenesis meeting looks something like this:

Introduction to the workshop;

Introduction to the Ontology for Biomedical Investigations (OBI) together with the devices and function branch; Overview of the rules for engagement with OBI;

• Task one: Familiarisation with Protege through placement of some devices and functions within the pre-existing structure;

• Task two: Development of attendee's own lists of devices and concomittant functions;

• Task three: Placement of personal lists in OBI together with metadata required by OBI;

• Task four: Development of object properties and placement of restrictions for devices and functions in OBI;

• Task five: Addition of defined classes over "normalised" devices tree.

• Task six: Addition of devices and functions to OBI at will from participants;

Feedback and discussion.

Rules of engagement:

• We follow OBI rules and conventions, that is, naming conventions, namespaces, etc. etc.

• The methodology used by OBI is not for debate--it is the way it is; we are, however, at liberty to provide feedback at the end of the workshop.

• We use OBO naming conventions ( http://obofoundry.org/wiki/index.php/Naming) for labels and semantic free identifiers for classes.

• We take the partially formed merged OBI owl file as starting material.

• We follow the normalisation methodology.

1.1. Dates/Time

December 15 and 16

1.2. Meeting Location

European Bioinformatics Institute (EBI), Wellcome Trust Genome Campus, Hinxton, CB10 1SD, Cambridge

IT Training room

1.3. Accommodation (in Cambridge)

• Travelodge, http://www.travelodge.co.uk/search_and_book/hotel_overview.php?hotel_id=255

• The Lensfield Hotel, http://www.lensfieldhotel.co.uk/

People wanting rooms should have booked these individually for one or both nights using their own credit cards before November 30th. Car parking is available both at the hotel (at least the first one) and the EBI.

1.4. Costs

The Ontogenesis Network will cover the costs of both workshop days, one nights accommodation and reasonable travel for Ontogenesis Network Members and speakers.

2. Programme of speakers and presentations

We will use Collaborative Protege ( http://protege.stanford.edu/doc/collab-protege/#UsersGuide) to extend the OBI device/instrument branch collaboratively, so there will only be a few introductory talks on OBI, the Collaborativ Protege tool and the whole setup. We'll provide a running and consistent OWL-DL version of OBI. Reading the OBI developers pages ( https://wiki.cbil.upenn.edu/obiwiki/index.php?title=HomePage) and looking at the ontology and its BFO top level classes would be a good idea. For those of you who need a revision of normalisation, the paper is at the end of this note. The primary aim is "delevop a small ontology collaboratively" (and later for a few of us to analyse and 'evaluate' issues that have been tracked by audio/video recording.

3. Attendees

Robert Stevens
James Malone
Daniel Schober
Frank Gibson
Kemian Dang
Christopher Brewster
Simon Jupp
Wes Sharrocks
Jennifer Clark
Nick Drummond
Kearon Mcnicol

4. Outcome

A wiki page we used to collect minutes of the meeting.

5. References

[1] Alan L. Rector: Modularisation of domain ontologies implemented in description logics and related formalisms including OWL. K-CAP 2003: 121-128. http://www.cs.man.ac.uk/~rector/papers/rector-modularisation-kcap-2003-distrib.pdf.

[2] Alan L. Rector, Chris Wroe, Jeremy Rogers, Angus Roberts: Untangling taxonomies and relationships: personal and practical problems in loosely coupled development of large ontologies. K-CAP 2001: 139-146. http://www.cs.man.ac.uk/~rector/papers/rector-k-cap-untangling-taxonomies-web.pdf.

6. Organisers

Robert Stevens ( robert.stevens@manchester.ac.uk), Daniel Schober ( schober@ebi.ac.uk) and James Malone ( malone@ebi.ac.uk)

---

-- DanielSchober 2008-01-10 16:27:51

---