UCLA Research on Information Management for Bioinformatics

This research initially focused on construction of a graph database for bioinformatics — a database system allowing storage and retrieval of labeled, directed graphs. Intuitively, graph databases seem well-suited for applications like bioinformatics that involve complex, unstructured information in a fundamental way. However, it is difficult to translate graph data management research into bioinformatics practice. Two deep issues arise: support for scalability, and support for ad hoc analysis. Traditional graph theoretic and database management approaches have not been able to resolve these issues in a satisfactory way. Currently, we are working on the following issues: Partial Order Alignment Alignment Indexing Bioinformatics-in-the-Large Graph Warehouses These issues reflect a data mining approach toward bioinformatics in which an initial information extraction phase focuses on joining information from multiple sources, using sequence alignment and other specialized indexes for join access paths, and a subsequent exploration phase emphasizes large-scale automated analysis. These issues are discussed in turn below.