NIST logo

Publication Citation: New Concepts for Building Vocabulary for Cell Image Ontologies

NIST Authors in Bold

Author(s): Anne L. Plant; John T. Elliott; Talapady N. Bhat;
Title: New Concepts for Building Vocabulary for Cell Image Ontologies
Published: December 21, 2011
Abstract: Background: We present an approach to cell image databasing that is compatible with searching across a global federation of independent image databases. The variety of biological experiments and data that practitioners would like to access and share is expanding rapidly. Because of the limitations posed by ontology schema, we suggest the use of ,root‰ terms and data tables as an alternative way to organize metadata so that different databases can be searched simultaneously. This approach makes it possible to easily add new terms locally within a hierarchical framework, and generate semantic queries on demand. Results: Using these concepts, a prototype database of more than 2500 images of cells and benchmark materials was compiled. A logical hierarchical structure is used to organize 163 metadata terms that describe experimental details about cell assays at a high level of granularity, including many details about cell culture and handling. An ontological file-naming scheme unambiguously identifies image and protocol files associated with each metadata value. Image files of interest can be retrieved by choosing one or more relevant metadata values as search terms. Metadata for any dataset can be compared with metadata of another dataset through a logical operation that returns metadata values that are non-identical between the data sets. Conclusions: This approach to organizing cell image data will make it easier to build and search federated databases. This ability will increase the usefulness of image data by permitting independent analysis, combination of results from different types of experiments, and elucidation of experimental differences that result in different observations.
Citation: BMC Bioinformatics
Volume: 12
Keywords: cell image; database; bioinformatics; semantic web; ontology; metadata; high content screening
Research Areas: Bioscience & Health
PDF version: PDF Document Click here to retrieve PDF version of paper (147KB)