> Home : Integration: Anatomy Case Study

Introduction:

To gain a better understanding of the issues involved in integrating mouse and human anatomy ontologies, a case study focussing on lung and branchial arch in mouse, and the corresponding tissues in human, is planned. This study takes a "vertical slice" through the target ontologies (listed below) and is intended to be complementary to an SOFG anatomy workshop (planned for early 2004) which aims at integrating mouse and human anatomy at a broad, but high and relatively shallow level of the ontologies (a "horizontal slice"). The case study has two objectives: 1) to specify the non-technical requirements in terms of the amount and level of anatomy knowledge that needs to be included in the ontology, and the necessary reasoning capabilities that should be associated with that knowledge, and 2) to write a proposal describing computational techniques that will adequately support these requirements.

Target Ontologies:

The suggested initial selection of target ontologies are:

If you would like any other ontology to be included in the initial analysis please let us know by 15 November 2003.

Target Tissue Systems:

The proposal is that the branchial arches (BA) and their derivatives are used as material for the case study - that is, to study in-depth a 'vertical' slice of the ontology. To include all the arches would probably be too much; the scale of the analysis can be cut down by focusing on just one or two arches. The benefit of being able to make this adjustment is that each arch contributes something different to the adult structure. The case for choosing the BA is as follows:

  • They are relatively simple structures that give rise to a restricted set of adult anatomical components each of which is reasonably complex. The several structures derived from even one BA are sufficiently different from one another to represent diverse issues for the ontology. For example, they include structures that may be considered as parts of different organ systems (cardio-vascular, hormonal, digestive, immune, nervous systems, etc.).
  • They present the issues associated with repeated units which have some common and some specific structures.
  • They present the common problems encountered with contiguous anatomical structures which, though distinct, are composed partly of tissues that are continuous from one structure to another (e.g. the epithelial and endothelial coverings of the arches).
  • They present the problems associated with left/right paired structures. In some instances, vasculature associated with left and right arches differs in its developmental fate.
  • They represent the issues encountered with transient developmental structures.
  • BA IV is significantly different in human and mouse (parathyroid IV, a derivative of BA IV, is present in human and absent from mouse; however, in both human and mouse a ultimobranchial body is formed from branchial pouch IV)

We also propose to use lung as material for the case study. The lung presents the general issues connected with an anatomical ontology without some of the more subtle problems presented by the more complex case of the branchial arches. It has the particular advantage of being a branched structure with a 'natural' form that fits almost ideally with the conceptual model of a tree. Also, there are differences between mouse and human in the numbers of lobes that will give us the opportunity to articulate simple differences between these organisms.

Planned Tasks:

Tasks: Outcome:
1. Description of relevant Sections of Target Ontologies:

For each target ontology, all relevant information regarding lung and branchial arch (or equivalent) needs to be described.
 Summary Report

(to be posted here once available)
2. Requirements Analysis (User Questionaire):

A questionaire has been produced to help identify user requirements with respect to mouse/human anatomy ontology integration. The questionaire will be sent out to all members of the SOFG mailing list, but will also be downloadable from here shortly. We would like to encourage all interested parties to fill in this questionaire to make sure that the requirements specification that will result from it is as inclusive as possible. (Note: details of individual submissions will not be made public, but appropriately summarised in the requirements specification.)
 Requirements Specification

(to be posted here once available)
3. Knowledge Representation and Reasoning Techniques:

Based on the Requirements Specification of Step 2 and the outcome of the SOFG anatomy workshop, appropriate computational techniques for integrating mouse and human anatomy are to be identified. The objective is to arrive at either a recommended form of representation for mouse and human anatomy ontologies and/or a set of minimal representation requirements which will enable the integration of such ontologies.
 Computational Specification

(to be posted here once available)

General Comments:

This case study is part of the wider SOFG efforts described on this Web site and, hence, is intended to be an effort by the community for the community. As such we would very much welcome input from all interested parties.

We would like to acknowledge the important contributions made by Alan Rector when he visited the HGU on 19th August 2003 and by the participants of a short teleconference session the following day (Richard Baldock, Albert Burger, Duncan Davidson, Alan Rector, Cornelius Ross and Mark Musen).

Contact:

This case study and the associated anatomy workshop are organised by Duncan Davidson and Albert Burger (Human Genetics Unit, MRC, Edinburgh). For more information, comments and questions, please contact Albert Burger.