Meeting with Deep Fin and other fishy folk

Phenoscapers met with Deep Fin and several associated fish groups in late February for ontology development, curation, web interface development and outreach. We had lots of good interactions with the 30+ ichthyologists attending. Read below for the details.

Phenoscape report from the DECAP meeting (Feb. 27-March 1, 2009)
(Paula Mabee, Wasila Dahdul, Jim Balhoff, Peter Midford)

Phenoscape met in association with the Deep Fin Research Coordination Network, Euteleost Tree of Life (EToL), Cypriniform Tree of Life (CToL), and All Catfish Species Inventory (ACSI), i.e. “DECAP”, in Kansas City February 27-March 1 to discuss “Updates, Resources, and Plans for the Phylogeny of Fishes”. The DeepFin network’s aim is to “build an open and diverse organization, to foster productive collaborations, to develop cyberinfrastructure, and to promote education on fish biodiversity and evolution”, and Phenoscape’s focus on providing an informatics connection among morphological and genetic data types is consistent with this mission.
Phenoscape organized a workshop a day before DECAP to use the gathering of ichthyologists to provide expert input to the Teleost Anatomy Ontology (TAO), to introduce them to the system, and to get feedback on the developing web interface. In attendance were Phenoscapers (Paula Mabee, Phenoscape PI; Wasila Dahdul, Anatomy ontology curator; Jim Balhoff, software and web-interface development; Peter Midford, Taxonomy ontology, liason with Catalog of Fishes), ichthyologists (Gloria Arratia, Terry Grande, Eric Hilton, John Lundberg, Rick Mayden, Mark Sabaj, Ed Wiley) and two graduate students (Jeff Engeman and Sarah Rages).
After introductions, Wasila gave participants a refresher/introduction to the Teleost Anatomy Ontology (TAO). She focused on the form of genus-differentia definitions that are required for ontology terms in preparation for our ontology development session. We sent spreadsheets of terms and other ontology data from two skeletal areas, the gill arches and the lateral line system, to focus on refining term definitions and relationships. These terms are required for curation of systematic characters, and some had been requested by individual curators via the term tracker over the past few months. Jackie Webb provided significant input to the lateral line system terms, and we thank her for this. Our goal was to obtain expert community review of these, focusing on consistency across terms and form and generality of definition.
We broke out into two groups (gill arches, lateral line system), and terms were evaluated one at a time. Additional synonyms and relationships were suggested by the group, definitions were refined and comments added. Some of the terms and definitions that were inherited from the zebrafish anatomical ontology (the TAO was cloned from the Zebrafish Anatomy Ontology in September 2007) required more general definitions. For example the definition for “hyomandibula” read “The hyomandibula is the large, dorsal-most member of the hyoid arch. It begins ossifying in the dorsal edge of the hyosymplectic cartilage near the hyomandibular foramen (4.6 mm NL). Ossification spreads through the cartilage, and sheets of membrane bone form off the cartilage model anteriorly and posteriorly (6.6 mm). In the adult, the hyomandibula has cartilage-capped anterior and posterior articulating heads that meet the sphenotic-prootic fossa and the pterotic fossa, respectively, in synovial joints. A posterior knob of bone develops, also capped in cartilage, and it forms a synovial joint with the opercle.” We recognized that the developmental information was not appropriate for the TAO because it applies to only zebrafish and internally suggested a new definition: “Endochondral bone that is bilaterally paired and articulates with the neurocranium dorsally, the opercle posteriorly, and the symplectic anteroventrally.” The experts, however, pointed out that not all teleosts have a symplectic, and thus a better formal definition is: “Endochondral bone that is bilaterally paired and articulates with the neurocranium dorsally and the opercle posteriorly.” Generalizations that don’t apply to all teleosts but which might be helpful to ontology users were written for the comments field in the ontology. For hyomandibula we added the comment: “In most teleosts, there is a symplectic, and the hyomandibula articulates with it anteroventrally.” Changes were made to these terms in the TAO by Wasila in the weeks following the meeting and these can be viewed using, e.g., the Bioportal browser (bioportal.org). This was a highly successful way to work through groups of terms, and we plan to do more group ontology development for other skeletal areas at the Joint Meeting of Ichthyologists and Herpetologists in Portland in July 2009.
In the afternoon, we focused the group on data curation, and Jim Balhoff and Peter Midford worked one-on-one with individuals in relation to the web interface and Teleost Taxonomy Ontology, respectively.
Jim Balhoff gave us an overview of the web user interface, and over the course of the afternoon and throughout the meeting, various participants from all of the projects gave him one-on-one feedback on the developing web user interface. Participants who provided feedback on the web interface mockups included Bill Eschmeyer, Sarah Rages, Rick Mayden, Ed Wiley, Eric Hilton, Gloria Arratia, Mark Sabaj-Perez, and Greg Riccardi, from Morphbank. The participants were led through screenshots of a preliminary web interface and suggested changes that would make the interface more usable for them.
During the afternoon, Peter Midford worked with curators on taxonomic issues. Suggested changes (additional synonyms, valid taxa) were made ‘live’ to the Teleost Taxonomy Ontology. Bill Eschmeyer and Stan Blum discussed future updates from CoF and Stan’s interest in the synonym extractor in TTO update with him. Peter also discussed CoF and taxonomy issues with Nicolas Bailly from FishBase. FishBase also imports taxonomy information from CoF.
Concurrently we focused on data curation. Curators were paired with participants who were unfamiliar with the Phenex, our curatorial tool, and they annotated data from papers that were either written by them or in their specific domain of expertise. We also “pre-curated” several papers of interest to the EToL group in order to introduce them to the project.
Paula gave an after dinner talk to the DECAP group, where she introduced Phenoscape and talked about the utility of the database in connecting morphological systematic studies to other data types.
The second day (Saturday) began with a morning of status reports from the various groups constituting DECAP. As Paula had already spoken about Phenoscape the previous evening, Jim and Peter filled Phenoscape’s morning slot with overviews of the informatics component and the TTO respectively. These overviews introduced these topics to DECAP participants who were not involved in Friday’s curation.
Saturday afternoon consisted of breakout groups, one of which was focused on taxonomy and name databasing. The discussion suggests that the TTO may have a role in the larger ichthyological community, as TAO already does. There was also a group focused on future work and a big taxonomy discussion to finish the afternoon.
The remaining two days of the meeting were spent in discussions with ichthyologists on Tree of Life topics, taxonomy and name databases, and issues concerning representing our knowledge about phylogenetic relationships of fishes. All of these topics relate to the synthesis that Phenoscape is trying to achieve, and the conference was of great value to us all.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: