Nuclear Receptor Signaling Atlas (www.nursa.org): hyperlinking the nuclear receptor signaling community

Abstract

The nuclear receptor signaling (NRS) field has generated a substantial body of information on nuclear receptors, their ligands and coregulators, with the ultimate goal of constructing coherent models of the biological and clinical significance of these molecules. As a component of the Nuclear Receptor Signaling Atlas (NURSA)--the development of a functional atlas of nuclear receptor biology--the NURSA Bioinformatics Resource is developing a strategy to organize and integrate legacy and future information on these molecules in a single web-based resource (www.nursa.org). This entails parallel efforts of (i) developing an appropriate software framework for handling datasets from NURSA laboratories and (ii) designing strategies for the curation and presentation of public data relevant to NRS. To illustrate our approach, we have described here in detail the development of a web-based interface for the NURSA quantitative PCR nuclear receptor expression dataset, incorporating bioinformatics analysis which provides novel perspectives on functional relationships between these molecules. We anticipate that the free and open access of the community to a platform for data mining and hypothesis generation strategies will be a significant contribution to the progress of research in this field.

Figure 1

Data sources and information exchange of the NURSA information solution. Emphasis is given to the NURSA databases, the NURSA website organization and the information exchange between the NURSA databases and website, and with external databases. BCM, Baylor College of Medicine; SIBS, Salk Institute for Biological Studies; UTSWD, University of Texas Southwestern at Dallas; UPMS, University of Pennsylvania Medical School; DUMC, Duke University Medical Center; CHCCC, City of Hope Comprehensive Cancer Center; VAI, Van Andel Institute; RDBMS, relational database management system.

Figure 2

Hyperlinking the research community. Screenshots taken from (status July 2005) indicating efficient, intuitive and intelligent browsing. Curated primary data (large screen) is cross-linked to far-reaching generic information: basic data housed within the NURSA domain (top right) and hyperlinked content of an external database (PDB in this example, bottom right). Primary NURSA data shown are peptides co-immunoprecipitated by immobilized estrogen receptor (ESR1, NR3A1)–coactivator CAPER (RNPC2) in HeLa cells (12).

Figure 3

Data mining efforts for integrated solutions. Combined correlation matrix (CM) for the Q-PCR anatomical expression profiling dataset for mouse nuclear receptors along with the results of cluster analysis (CA) and data integration (SD) allow for ‘at-a-glance’ interpretations of a comprehensive dataset as well as rapid data access. The later is illustrated by a screenshot (SS), which shows a side-by-side representation of the expression profiles of ER-alpha in 39 tissues of 129x1/SvJ and C57Bl/6 mice (see text for details).