Database Design and Development

ArrayDB 2.1

Microarray technology is providing the means to do large-scale or whole-genome expression studies. While this will undoubtedly push the envelope of our current understanding of gene expression and regulation, it poses new challenges in data tracking and analysis. To address this issue, the Array Informatics Team, under the direction of Dr. Andreas Baxevanis, has developed an array informatics system which integrates data management and analysis.

We have developed a Sybase database schema to manage a wide variety of data. This relational database stores information including image data, experiment data (printing arrays, probes), and clones (IMAGE Clone ID, Title, UniGene cluster). A commercial middleware product, Websql, is utilized to provide real-time interactive queries of the database through a convenient and widely accessible Web interface. The Web interface is provided as both standard HTML Web pages and, where a greater degree of interaction is needed, Java applets have been implemented.

Those accessing the database can obtain information about clones, including performing homology searches against the 15K set. Furthermore, the systems provide the means to analyze experiment data sets individually or in groups. A Java applet provides the single experiment data analysis interface, where the image data taken from the database is used to draw a histogram of the ratio values with confidence limits. A list of clones falling outside the confidence limits (being therefore statistically significant) is linked to value-added information including the UniGene cluster, chromosomal location, and metabolic pathway information. System users can also identify patterns in high or low ratios across several experiments.

Informatics Design