(prepared by Eck Doerry)

Development of a complete and uniform anatomical dictionary for the zebrafish is vital to the success of zebrafish science. Specifically, the dictionary is necessary for:

• Effective data dissemination and informatics. Reference to anatomical structures is the central to all phenotypic description. For example, in situ expression patterns are described in terms of the structures in which expression is observed; mutations are described in terms of the anatomical structure that are affected. In this way, anatomical structure is the common ground on which expressed genotype meets mutational phenotype, yielding powerful insight into gene function. A concise and consistent anatomical atlas is necessary to support this correlation, particularly in an automated or computer-aided fashion.
• A reference framework. By defining what anatomical structures exist, the dictionary establishes an ontological framework for a variety of reference resources, including a concise staging series (i.e., defined by development of structures, not time), an atlas of reference images indexed by anatomical structure and developmental age, and annotated 3-D reconstructions.
• Interoperability. A stable, concise anatomical atlas is the key to leveraging work done in other species, particularly through semi-automated mapping between zebrafish data (stored in ZFIN) and data stored in other species’ databases (e.g. MGI,Flybase). Given dictionaries for two species, mapping relationships between analogous structures (e.g. fins in fish=legs in flies) can be developed; searches based on mutations to zebrafish fins could yield genes expressed in drosophila legs.

Although the need for an anatomical dictionary has been recognized for some time, development been slow. Bill Trevarrow made a tremendous start in 1996, but progress stalled due to anatomical quandaries and lack of time to resolve them. In addition, several pragmatic efforts have yielded rudimentary dictionaries to characterize specific sets of in situ or mutant data. It is vital that these efforts be coordinated before they conflict and/or diverge too far; untangling the chaos post hoc would be a nightmare.

To make a start at coordinating development of the dictionary, an informal breakfast meeting of interested parties was arranged on May 11,1999, at the annual NIH-sponsored workshop on zebrafish genomic and genetic tools. Although not all interested parties were able to attend, this provided a convenient, low-cost, low-commitment opportunity for introductions and quick overviews of projects, perspectives on the dictionary, and planning for future development.

Others who have expressed interest but could not attend:

• Jonathan Bard (Edinburgh). Jonathan was the driving force behind the development of the mouse anatomical dictionary. He has recently begun doing the same for the zebrafish. A draft of a zebrafish atlas Jonathan has developed was circulated at the meeting.
• Fons Verbeek (Utrecht). Fons is developing annotated 3-D reconstructions (from serial sections) of zebrafish embryos. When fully annotated, it is expected that this effort will eventually yield a definitive 3-D atlas for the zebrafish.

Predictably, everyone in attendance agreed that developing the anatomical dictionary is a high priority. The need is so urgent, in fact, that we find ourselves in the uncomfortable position of having to develop the dictionary, while at the same time using it to characterize data being produced today. The following points sketch out specific conclusions:

• There was overall agreement and commitment to work collaboratively on the dictionary, making every effort to synchronize development.
• No one person or source of expertise adequate for developing the dictionary exists. Thus, the dictionary should be developed in a distributed fashion, drawing on expertise wherever it exists. For example, experts on CNS structure might establish a subgroup to discuss and develop this aspect of the dictionary; similar subgroups would form around other specialties.
• We will establish a central site as the clearinghouse for information related to the dictionary; the latest "version" of the dictionary will always be available here. Although rudimentary at the outset, this site might eventually provide tools and resources to support the distributed, collaborative development of the dictionary. In particular, the site will provide a mailing list for the working group, and will archive the critical discussion on the list as a way for preserving development rationale. ZFIN will host this site, at least initially.
• We will pursue funding for regular meetings (at least annually) of the working group and for 1-2 dedicated staff to organize the effort, solicit input from experts where needed, and curate contributions. development of the dictionary.
• The first meeting is scheduled for December 10, 1999. Reiko Toyama and Monte Westerfield are co-organizers.