Mouse anatomy ontologies: enhancements and tools for exploring and integrating biomedical data

Abstract

Mouse anatomy ontologies provide standard nomenclature for describing normal and mutant mouse anatomy, and are essential for the description and integration of data directly related to anatomy such as gene expression patterns. Building on our previous work on anatomical ontologies for the embryonic and adult mouse, we have recently developed a new and substantially revised anatomical ontology covering all life stages of the mouse. Anatomical terms are organized in complex hierarchies enabling multiple relationships between terms. Tissue classification as well as partonomic, developmental, and other types of relationships can be represented. Hierarchies for specific developmental stages can also be derived. The ontology forms the core of the eMouse Atlas Project (EMAP) and is used extensively for annotating and integrating gene expression patterns and other data by the Gene Expression Database (GXD), the eMouse Atlas of Gene Expression (EMAGE) and other database resources. Here we illustrate the evolution of the developmental and adult mouse anatomical ontologies toward one combined system. We report on recent ontology enhancements, describe the current status, and discuss future plans for mouse anatomy ontology development and application in integrating data resources.

Fig. 1

Mouse anatomy ontologies enable standardized description of mouse anatomy for data from different sources. a Histological sections from The Atlas of Mouse Development, with anatomical structures identified by Kaufman, provided the initial list of tissues for the developmental mouse anatomy ontology. Ontology terms have since been used to label structures included in an online version of this resource. b Ontology terms have also been used by the eMouse Atlas to identify anatomical domains in 2D and 3D models of mouse embryos. c GXD uses mouse anatomy ontology terms for textual description of gene expression patterns. Consistent use of a common standardized anatomy nomenclature enables and facilitates data integration between resources. Cover image A reprinted from Kaufman (1994) with permission from Elsevier

Fig. 2

Anatomy ontologies present anatomical terms within a hierarchically organized format, and describe relationships between the anatomical structures represented. a The original anatomy ontology for the developing mouse was designed as a simple uniparental hierarchy, exclusively using part-of relationships, with separate trees for the different Theiler stages of development. b The EMAPA ontology now provides a unified representation of mouse anatomy for all stages of development, and supports multiple parentage and relationship types. c Stage-specific EMAPS hierarchies containing only those terms relevant for a specific stage are derived from the EMAPA using stage-range information for each term contained within the EMAPA file

Fig. 3

The Mouse Developmental Anatomy Browser enables text-based searches as well as tree-based browsing of the ontology. a The Anatomy Search tool enables searching for anatomy terms using structure names and synonyms. b and c Anatomical Term Detail and hierarchical Tree Views are displayed for both “abstract” EMAPA (b) and stage-specific EMAPS (c) representations of the anatomy ontology, with the ability to freely toggle between them. Direct links are provided to GXD expression result summaries for the selected term and its hierarchical children

Fig. 4

GXD results summaries provide separate tabs for Genes, Assays, Assay Results, Images, and matrix views which present expression data at different levels of detail. a The Assay Results tab provides a report of all annotated results relevant to the selected anatomical structure and/or other query parameters, and features links to corresponding Assay Detail pages. b The Tissue x Stage Matrix presents a combined spatial and temporal overview for a set of expression results. c The Tissue × Gene Matrix enables a comparison of expression patterns for all genes in a results set. Summaries can be iteratively refined by modifying the search or by applying various data filtering options